, not C:/cygwin/home/Administrator. This means that environment variables are not exported.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7069","12/06/2019 22:01:44",2,"Parameterize New Relic spruce browser plugin ""EVG-7028 parameterized New Relic for the legacy Evergreen UI. We should also parameterize it for Spruce. Unlike the Evergreen UI, where something can be stored in the admin interface, this will need to passed in to the Spruce buildsystem.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7070","12/09/2019 14:15:04",0,"Fix static host reprovisioning to legacy and tune job enqueueing ""Fix various issues: * If we're converting a static host to legacy provisioning, it has to be able to use a Jasper client to stop the agent monitor. * The host event page has the wrong event type name. * Jobs should be rounded to the nearest minute for uniqueness to allow provisioning to be changed multiple times per hour. Otherwise, the host will have to wait until the next deploy to be reprovisioned. * Needs new agent/monitor should be checked only after the job is locked.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8674","12/09/2019 19:36:22",1,"Simplify RunCMD ""This was suggested by the gocognit linter and I do think that the method is excessively complicated and should be broken into some helper functions.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7074","12/09/2019 22:44:21",1,"Change the TAG API to be part of the update project API (or at least similar to it) ""There are a few reasons to ask for this: 1. someone may want to add/remove more than a single tag. 2. someone may want to change from one tag to another (would require two calls instead of a single call with {tags: ['new'], tags_to_remove: [""""old""""]} 3. most likely someone may want to already make some other changes to a project in addition to changing its tags. This way a single update call can be used to make all the required changes.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7075","12/10/2019 14:40:05",1,"evg-lint needs to handle nolint ""It must be able to handle nolint comment directives as part of migration to golangci-lint instead of gometalinter. golangci-lint does not have easy integration with custom linters because it does not support running linter binaries.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7076","12/10/2019 14:43:12",0,"Host that does not come up can cause all host provisioning to fail ""At 6:07pm EST on 12/09/19, a user requested spawn host i-0489a525d49f1af12. This spawn host hit the [host not ready for setup|https://github.com/evergreen-ci/evergreen/blob/f50768d851142375ee202210e4f6424128199c79/units/host_status.go#L130] case from 6:08pm to 7:10pm, suggesting either that it's possible for a host to get stuck in some pending status or that there is some failed status that Evergreen does not handle correctly. After the hour passed, AWS no longer returned data about that host. This caused Evergreen to [return early|https://github.com/evergreen-ci/evergreen/blob/f50768d851142375ee202210e4f6424128199c79/cloud/ec2.go#L724] from GetInstanceStatuses, because that host was no longer included in the DescribeInstances response. This in turn blocked on host provisioning that went through that DescribeInstances code path. Since fleet is its own provider, most task hosts did come up, but hosts spawned by users, hosts spawned by host.create, and on-demand hosts (i.e., perf) did not. We fixed this in production by terminating all on-demand hosts older than 1 hour. One possible fix for this is to terminate a host in the database if it is not included in the DescribeInstances response. Another is to figure out the additional case we are failing to handle in setCloudHostStatus.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7080","12/10/2019 21:50:03",2,"Setup Failure: test-cloud on Ubuntu 16.04 [Evergreen @ b5d79830] ""h2. [test-cloud failed on Ubuntu 16.04|https://evergreen.mongodb.com/task/evergreen_ubuntu1604_test_cloud_b5d798308e84a22c9f8aece21c506c3fdd0037e7_19_12_10_20_37_30/1] Host: [ec2-3-91-38-130.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-0526f1bd337f73f84] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/b5d798308e84a22c9f8aece21c506c3fdd0037e7]: EVG-6882 prevent triggered builds for disabled projects (#2952) | 10 Dec 19 20:37 UTC Evergreen Subscription: ; Evergreen Event: [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_ubuntu1604_test_cloud_b5d798308e84a22c9f8aece21c506c3fdd0037e7_19_12_10_20_37_30/1?type=T ] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-8836","12/11/2019 13:42:58",1,"Deduplicate Jasper remote tests and do not skip tests ""There are many copy-pasted tests in the remote package which are named slightly differently but test the exact same things. They should be unified under the common remote tests. Fix this test: [https://evergreen.mongodb.com/task/jasper_windows_test_remote_61a695020101f18236583d806735d4ea1ccfd1fe_19_12_11_15_47_40] The tests that are being skipped should most likely not be skipped and should be fixed.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7100","12/11/2019 16:10:09",2,"Remove TaskOrdering from code ""ubuntu1604-test is running the new tunable scheduler. I observed that a patch was at the top of the queue https://evergreen.mongodb.com/version/5df10c3030661525ab8d4acf, followed by all the tasks for this mainline commit https://evergreen.mongodb.com/version/mongodb_mongo_v3.6_919b7c1ce02104a22a06b14016d152440f66161d, followed by all the tasks in other patches https://evergreen.mongodb.com/version/5df112e2a4cf47113ff144d7. This is unexpected because I would expect """"interleave"""", which is what, to interleave patch and mainline commits.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7085","12/11/2019 19:20:49",2,"Toggle UseParserProject to true ""This cannot be done until EVG-7077. Test extensively in staging. Once this is done and in for a safe amount of time we can revisit EVG-5313. ""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8870","12/11/2019 20:00:13",2,"fix model.TestGetDownloadWindows ""I've added a skip""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8459","12/11/2019 20:01:33",1,"Task mapping creation should increment source_file_seen_count even when no tasks flip ""For this file in src/evergreen/__init__.py file in evergreen.py, it has changed in 14 commits since this commit on May 4: https://github.com/evergreen-ci/evergreen.py/commit/61ee0d619d4adb31c84abc02dd9d52d41ae45f63. Thus, when I run the following command in the selected-tests repo, the task mapping that is created should have a source_file_seen_count of 14. Currently the task mapping has a source_file_seen_count of 2 (see below). This is because we are only incrementing source_file_seen_count when tasks in Evergreen have flipped for that commit. Instead, we want to increment source_file_seen_count regardless of whether tasks have flipped. """," task-mappings create evergreen.py --source-file-regex """"^src/evergreen/__init__.py"""" --after 2019-05-04T00:00:12 --output-file output.txt [ { """"source_file"""": """"src/evergreen/__init__.py"""", """"project"""": """"evergreen.py"""", """"repo"""": """"evergreen.py"""", """"branch"""": """"master"""", """"source_file_seen_count"""": 2, """"tasks"""": [ { """"name"""": """"unit_tests"""", """"variant"""": """"ubuntu1604"""", """"flip_count"""": 2 } ] } ] ",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7088","12/11/2019 22:48:13",2,"Slack notification says succeeded but task actually failed ""This PR patch was cancelled because I pushed another commit, but the slack notification says that it succeeded: [https://evergreen.mongodb.com/version/5df17050c9ec444c0b932df0] Subscription: 5bd86621e3c331358aa373f3 Event: 5df1718e3066155171dbbe2a""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7089","12/12/2019 15:17:14",0,"Copy teardown script to host ""This might cause user data to hit its limit (it has a 16kb limit) since the teardown script can be arbitrarily large.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7090","12/12/2019 18:04:59",1,"pip failure on windows-64-vs2017-compile server ""Our automation agent MSI builds have been failing since yesterday (12/11/2019) on windows-64-vs2017-compile. Here is a link to the failed build: https://evergreen.mongodb.com/task/cloud_automation_master_build_windows_msi_Automation_Agent_Generic_0a37709a3505599f828defd6e1df66fc0782671f_19_12_12_17_14_00 cc [~kimberly.tao] [~tomer.yakir] ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7092","12/12/2019 19:04:45",2,"Improve host creation throttling ""Since all task hosts are throttled together it's possible to starve out a single distro and tasks waiting on that distro will not run. Perhaps a round-robin over the distros would allow each distro to be serviced.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7093","12/12/2019 19:37:05",1,"Updating spawn hosts doesn't refresh the page ""Clicking the Update button doesn't refresh the spawn hosts page to show the new properties and the banner.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7095","12/13/2019 17:50:34",1,"Periodic builds include all tasks ""Periodic builds include all tasks, not just the ones that match the alias.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7097","12/13/2019 19:29:36",0,"Task directories not being cleaned up between tasks ""Windows agents are failing to clean up their task directories and leave behind files between tasks. Example errors in agent logs: {{agent-3424-14.log:[evergreen.agent] 2019/12/13 17:57:46 [p=critical]: Error removing working directory for the task: \data\mci\24b47f5b24b3be8ac822dea79146ea26: remove \data\mci\24b47f5b24b3be8ac822dea79146ea26\mms\server\build\java\classes\com\xgen\svc\mms\api\view\atlas_private\dataLake: The directory is not empty.}} {{agent-3424-16.log:[evergreen.agent] 2019/12/13 18:15:22 [p=critical]: Error removing working directory for the task: \data\mci\5d8f220a77d23a59aba3912c4a97d73f: remove \data\mci\5d8f220a77d23a59aba3912c4a97d73f\mms\server\build\java\classes\com\xgen\svc\mms\api\view\atlas_private\dataLake: The directory is not empty.}} {{agent-3424-18.log:[evergreen.agent] 2019/12/13 18:33:11 [p=critical]: Error removing working directory for the task: \data\mci\4b701a6ca0f7fbaac61a430895e05133: remove \data\mci\4b701a6ca0f7fbaac61a430895e05133\mms\server\build\java\classes\com\xgen\svc\mms\api\view\atlas_private\dataLake: The directory is not empty.}} There is also a related issue where a Windows agent is cleaning up a task directory but then stops communicating with the app server, causing the app server to attempt to deploy agents/agent monitors to the host when the agent is still running. This applies to both non-legacy and legacy provisioning on Windows machines. I checked on windows-64-vs2013-small and windows-64-vs2015-compile and both had this issue. Example of a legacy host (windows-64-2019) in production that had this issue: https://evergreen.mongodb.com/host/i-00807d0b61df6bdfc I quarantined some hosts in staging that seem to have this issue (using legacy provisioning). I'm unsure if they're the same issue but they have the same symptom of not being able to clean up the task directory: [https://evergreen-staging.corp.mongodb.com/host/i-035a2aab525bd8c77] [https://evergreen-staging.corp.mongodb.com/host/i-0218a3aa604dbfa88]""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7098","12/13/2019 19:49:50",1,"Add a notes field to distros ""We have created more custom distros which have very specific use cases. If you don't have a knowledge of the history of the distro, it's easy to change distro settings in a way that breaks implicit assumptions about how the distro will be used. We should add a notes field to the distro, visible in the UI, so that administrators can record facts about the distros.""","",0,0,0,1,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8973","12/13/2019 20:57:18",1,"make bson documents json marshalable (native) ""this shouldn't be extended json support.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8607","12/16/2019 18:53:56",2,"Cron backfiller for signal processing recalculation ""I'd like a schedule based backstop to recalculate change points over cedar's data. Such that, any stale time series get refreshed regularly. AC: * Recalculate change points for all cedar metrics with new data. Run every hour.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7101","12/17/2019 14:58:02",2,"Sys Perf trend charts in patchbuilds can misorder results ""In this [patchbuild|https://evergreen.mongodb.com/task/sys_perf_linux_standalone_audit_industry_benchmarks_patch_58abcf6577982367232a6c76d1ee277a4031ed3c_5dead5740ae6060c325daab9_19_12_06_22_26_35##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%252258abcf6577982367232a6c76d1ee277a4031ed3c%2522%257D%255D%257D], the output was rendered on the far left of the trend chart. The patchbuild was performed December 6th 2019 off of a recent commit. However, the visualization places it near patches from 2018. Placing the results for the patchbuild far away from related results makes it more difficult to visualize changes. This commit: !image-2019-12-17-09-55-56-482.png|thumbnail! The """"next"""" commit: !image-2019-12-17-09-56-28-389.png|thumbnail! ""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8468","12/17/2019 15:11:51",1,"Task mapping creation should set repo value to module's repo name for module-related task mappings ""When I create task mappings for the mongodb-mongo-master project and enterprise module, the task mappings that are generated for the enterprise module have their repo value set to the project repo rather than to the module repo. Command to create task mappings: Here is one of the task mappings created by the above command: This mapping corresponds to [this file|https://github.com/10gen/mongo-enterprise-modules/blob/master/src/queryable/SConscript] in the mongo-enterprise-modules repo. The """"repo"""" field says """"mongo"""" when it should say """"mongo-enterprise-modules"""". """," task-mappings create mongodb-mongo-master --source-file-regex """"^src"""" --after 2019-05-04T00:00:12 --source-file-regex """"^src"""" --module-name """"enterprise"""" --module-source-file-regex """"^src"""" --build-variant-regex """"^\!"""" --output-file output.txt { """"source_file"""" : """"src/queryable/blockstore/SConscript"""", """"project"""" : """"mongodb-mongo-master"""", """"repo"""" : """"mongo"""", """"branch"""" : """"master"""", """"source_file_seen_count"""" : 4, """"tasks"""" : [ { """"name"""" : """"aggregation_multiversion_fuzzer_1_enterprise-rhel-62-64-bit"""", """"variant"""" : """"enterprise-rhel-62-64-bit"""", """"flip_count"""" : 1 }, { """"name"""" : """"aggregation_multiversion_fuzzer"""", """"variant"""" : """"enterprise-rhel-62-64-bit"""", """"flip_count"""" : 1 }, { """"name"""" : """"aggregation_wildcard_fuzzer"""", """"variant"""" : """"enterprise-rhel-62-64-bit"""", """"flip_count"""" : 1 }, { """"name"""" : """"multiversion_auth"""", """"variant"""" : """"enterprise-rhel-62-64-bit"""", """"flip_count"""" : 1 }, { """"name"""" : """"rollback_fuzzer_clean_shutdowns_1_enterprise-rhel-62-64-bi t"""", """"variant"""" : """"enterprise-rhel-62-64-bit"""", """"flip_count"""" : 2 }] } ",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7104","12/17/2019 22:58:13",1,"Prevent using regular patch and set-module commands for commit queue patches ""Users should not be able to create a commit queue patch with the regular patch command. Also, if a module is added with the regular set-module command the message will be empty and the commit will fail. We should error if a user does this.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7105","12/17/2019 23:02:07",0,"Commit queue set-module won't use the description argument ""Since the message flag was removed (https://github.com/evergreen-ci/evergreen/commit/8f99d8f744f39051ee6fd5338ce977e15483b4a6) pulling the message flag from the context will get the empty string.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8391","12/18/2019 20:13:52",1,"Cedar: problem persisting argument adding sample #10002: collector is overfull ""After adding O(1000000) events closing the stream fails with an error similar to the following: problem persisting argument adding sample #10002: collector is overfull ""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8627","12/18/2019 22:13:57",1,"Move selected-tests to python 3.7 ""It recently came up that using Python 3.6 prohibits us from using annotations: https://github.com/mongodb/selected-tests/pull/30#discussion_r359564871 We should move our selected-tests deployment onto Python 3.7 by updating our Docker image.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7107","12/19/2019 19:31:25",2,"Design GQL schema ""Final version of the schema depends on getting final version of the mockups from product designer. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7108","12/19/2019 19:32:43",1,"Map schema to database models ""Design doc: [https://docs.google.com/document/d/1vvvPkIgjgFq4iFq239b37kjzQ-gcf2Xw1suc730SyWo/edit#heading=h.4jq6vrtk2brz]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7109","12/19/2019 19:34:08",2,"Handle authentication ""Design doc: [https://docs.google.com/document/d/1vvvPkIgjgFq4iFq239b37kjzQ-gcf2Xw1suc730SyWo/edit#heading=h.lazhk4iywuu2] Handling authentication in graphQL server should be done using middleware to inspect the cookie from incoming requests. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7111","12/19/2019 19:40:22",1,"Configure CORS in gql server ""Design doc: [https://docs.google.com/document/d/1vvvPkIgjgFq4iFq239b37kjzQ-gcf2Xw1suc730SyWo/edit#heading=h.gzmckhx4wevo] # Support queries from localhost when in dev or test environment # Support origin url of spruce""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7112","12/19/2019 19:41:26",1,"Disable introspection in production ""Design doc: [https://docs.google.com/document/d/1vvvPkIgjgFq4iFq239b37kjzQ-gcf2Xw1suc730SyWo/edit#heading=h.4va3kae6h7ev]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7113","12/19/2019 19:42:16",2,"Support pagination in queries ""This is the full stack feature for supporting pagination in GQL queries between the client and server. Apollo Client queries for paginated requests include a cursor and the GQL server returns paginated data based on that cursor. Design doc: [https://docs.google.com/document/d/1vvvPkIgjgFq4iFq239b37kjzQ-gcf2Xw1suc730SyWo/edit#heading=h.yqh5m6v86y8s]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7114","12/19/2019 19:48:41",2,"Dependency inject mock data connector for integration testing ""Design doc: [https://docs.google.com/document/d/1vvvPkIgjgFq4iFq239b37kjzQ-gcf2Xw1suc730SyWo/edit#heading=h.rpf8yyiq0cg5] Support integration tests for the server by instantiating a test gql server with the mock data connector""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7115","12/19/2019 19:55:33",2,"Start gql server in evergreen behind feature flag ""Design doc: [https://docs.google.com/document/d/1vvvPkIgjgFq4iFq239b37kjzQ-gcf2Xw1suc730SyWo/edit#heading=h.l5hu22fsyoxi]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7118","12/20/2019 14:55:11",1,"Add styled lib to Spruce ""Adding the library and some default styles should be enough. The existing styles do not need to be replaced because most of Spruce UI (the patches page) will be deleted once we get the mockups for the new UI.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7120","12/20/2019 15:36:47",1,"Add distro feature to download legacy agent ""EVG-7119 adds the ability to build a legacy agent on an older Go. We should have a feature to opt a distro into deploying this older agent.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7121","12/21/2019 21:10:02",2,"Add created_date to collections that do not have it ""This would aid in dumping a limited set of data, to aid in creating local Evergreen for development, and also for DR to quickly get a limited set of data up. This does not apply to the logs or test results collections, which are large enough to be ignored for any local Evergreen or DR initial restore. Remember also add an index on this field.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8476","12/27/2019 18:45:19",1,"test selection should return data as json ""It looks like the test-selection service is returning most of the data as a string instead of json. This will be annoying to consumers of the service and should be changed. See [here|https://selected-tests.server-tig.prod.corp.mongodb.com/projects/evergreen.py/test-mappings?changed_files=src%2Fevergreen%2Futil.py&threshold=0.5] for an example. ---- As a test-selection consumer, I want all the data returned as json, So that I can easily parse it. ---- AC: * All endpoints return their data as json. * Tests are written to ensure this is the case.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7127","12/30/2019 16:32:20",1,"Duplicating a project creates empty redacted variables ""If you create a project by clicking """"New Project"""" and selecting """"Duplicate current project"""", the new project will include the keys of the redacted variables from the old project, but their values will be empty. This is confusing for users, as it appears that the redacted variables were copied successfully.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7130","01/02/2020 19:35:51",2,"Adding a tag fails before AWS gives us an instance ""Trying to add a tag to a spawnhost before we get an instance fails because we try to add tags to the instance and its resources before they actually exist. It's not made clear to the user what went wrong (the request eventually returns 500). We should prevent a user from attempting to set tags until the host is provisioning.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7132","01/06/2020 12:33:28",1,"create_time is being over ridden with the default ""The ordering in [getDefaultFiltering|https://github.com/evergreen-ci/evergreen/blob/060c9dac8bf27c860857275185843a0743659d61/public/static/app/perf-bb/SignalProcessingCtrl.js#L38-L45] of __Settings.perf.signalProcessing.persistentFiltering__ and __mandatoryDefaultFiltering__ needs to be swapped in the following extract: """," const getDefaultFiltering = function() { return _.extend( {}, secondaryDefaultFiltering, Settings.perf.signalProcessing.persistentFiltering, mandatoryDefaultFiltering ); }; ",0,0,1,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-8624","01/06/2020 16:01:18",2,"Integrate mypy ""As a DAG engineer, I want to know that functions and methods in the selected-tests repo have correct type annotations, so that fewer bugs are introduced to the code base. AC * mypy is run through pytest as part of CI""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8849","01/06/2020 16:05:03",2,"Set up Okta user manager ""Similar to the OAuth client for GitHub.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7138","01/06/2020 18:44:37",2,"Volume attach command times out ""Trying to attach a volume to a host from the CLI consistently times out and cancels the context so the volume never gets attached. Maybe we should return immediately and optionally subscribe for notifications, similar to what we do for spawning hosts from the CLI.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8884","01/06/2020 18:56:30",1,"Implement user manager logout ""This should clear the user in the database so that evergreen considers them logged out (i.e. delete their cached login tokens). Their Okta login status should be unchanged.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7139","01/06/2020 20:24:59",1,"Retry failed GQL requests ""Decide on how we should implement apollo-link-retry""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7140","01/06/2020 21:06:02",2,"Update local GQL schema from GitHub as a dev build step ""Conditionally update local GQL schema from evergreen GitHub repo as a dev build step""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7145","01/06/2020 22:28:12",1,"Document VS Code remote development setup and workflow ""Document VS Code remote development setup and workflow""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8562","01/06/2020 22:28:13",2,"Jasper managed daemon processes (auto-restart option on creation) ""Jasper managed daemon processes (auto-restart option on creation)""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7148","01/06/2020 22:28:13",1,"Start code-server on spawnhosts ""Start code-server on spawnhosts""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7149","01/06/2020 22:28:14",0,"Reverse-proxy code-server through Evergreen app servers ""Reverse-proxy code-server through Evergreen app servers""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7152","01/06/2020 22:28:15",0,"Admins can run a script on groups of hosts ""Admins can run a script on groups of hosts""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7151","01/06/2020 22:28:15",0,"Spawnhost option to mount a volume on a user’s home directory ""Spawnhost option to mount a volume on a user’s home directory""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8753","01/06/2020 23:12:26",2,"Switch to FastAPI framework ""As a DAG engineer, I would like to use FastAPI rather than Flask + flask-restplus to build selected-tests service, so that my endpoints have faster response times and so that coding in the application code is faster and easier to do. AC: * Use FastAPI * Endpoints maintain good documentation equivalent to current flask-restplus documentation * HTTP requests are logged to stdout so we can see them in Splunk""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7153","01/07/2020 13:41:06",2,"Add admin settings for Okta auth ""Requires client ID, client secret, issuer, reauthorization TTL, and user group name.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7154","01/07/2020 13:44:44",0,"Create admin setting to select auth method ""Currently, the auth method is chosen based on which auth fields in the admin settings are populated in {{LoadUserManager}}. It would be nice to be able to manually select one if multiple auth methods are supported.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8997","01/07/2020 14:35:13",0,"Handle reauthorization after given expiration time ""The current LDAP manager reauthorizes users after a TTL (configured to one hour right now) by checking the user's LDAP groups. The new Okta user manager should also do some kind of periodic checks against Okta. Checking the user group (in a similar way as LDAP is currentl ydone) requires using the user's access token, which expires after some amount of time. When the token is expired, we have to handle getting a new one. This could potentially be achieved by requesting refresh tokens. If the access token and refresh token are invalid, then we have to redo authentication to get more tokens. There's also the possibility that we can't use refresh tokens (i.e. due to security policies), so we might have the user go through silent reauthentication (without a login prompt). If we go this route, we have to somehow know where to redirect the user.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7156","01/07/2020 15:22:43",2,"Integrate Okta user manager ""Mostly this is passing the admin settings to the Okta user manager to create it. Additionally, we have to set the access and refresh tokens with the caching functions.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8792","01/07/2020 16:34:43",2,"add support for okta ldap to gimlet ""be able to support okta and legacy/standard ldap""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7159","01/07/2020 18:12:23",0,"Uneject Create React App ""our current version of create react app has been ejected. we should be running a version of create react app that has not been ejected. The unejected CRA should have: 1) the fractal file layout pattern 2) the same deploy steps as before w/ npm scripts 3) the GQLWrapper and tests 4) environment variable utils 5) New Relic plugin 6) Bugsnag integration 7) absolute imports instead of relative imports 8) same readme as before 9) same npm scripts to start mock & dev server ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7163","01/07/2020 20:20:26",2,"Parent hosts shouldn't be running agents ""https://evergreen.mongodb.com/host/i-07f82fbbe9fcfa38c https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Devergreen%20i-07f82fbbe9fcfa38c&display.page.search.mode=verbose&dispatch.sample_ratio=1&earliest=1578426540&latest=1578428377&sid=1578428433.4285417""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7166","01/07/2020 21:55:10",1,"Hide Never expire checkbox when there are 0 unexpirable hosts ""(On the spawn host page)""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7167","01/07/2020 21:57:41",0,"Send Parser Project over the wire instead of yaml ""For example, [https://github.com/evergreen-ci/evergreen/blob/master/service/rest_version.go#L310] converts ParserProject back into yaml, which won't be necessary when Parser project is done.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7168","01/08/2020 13:18:35",1,"Query linting ""All GraphQL queries on the client should be linted. This can be achieved by using the GQL schema.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7174","01/08/2020 18:39:04",0,"fix perf charts for stitch (BAAS) ""perf charts still not showing up for projects that do not have legacy perf data (such as stitch)""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8859","01/08/2020 18:42:15",1,"Implement and persist nonce and state ""For the authentication process, we have to get access/ID tokens so we can authenticate/authorize the user. To do so, we have to temporarily store two secrets (nonce and state) because it is required to get the access/ID tokens when the redirect URI callback is invoked. The nonce and state should have a TTL since they are not used after the user is authenticated.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7176","01/08/2020 22:30:47",2,"Automate migrating between instances ""Detach the old instance's home volume and attach it to the new host.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7177","01/09/2020 14:38:16",2,"Create model for storing required Okta auth information "" Once we've authenticated the user, we need to store each user's token (to find users in the database), access token (for checking groups), refresh token (for getting new access tokens without having to reauthenticate the user), TTL (to reauthorize the user periodically).""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7178","01/09/2020 15:37:39",1,"Updating a project without projectVars panics ""projectVars is dereferenced [here|https://github.com/evergreen-ci/evergreen/blob/f036124c01bd7de98e2b03e660636d08634a6fe9/service/project.go#L526] and if it isn't found [here|https://github.com/evergreen-ci/evergreen/blob/f036124c01bd7de98e2b03e660636d08634a6fe9/service/project.go#L458], projectVars will be nil. It's hard to know how a project could have been created without [also inserting projectVars|https://github.com/evergreen-ci/evergreen/blob/f036124c01bd7de98e2b03e660636d08634a6fe9/service/project.go#L636].""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7183","01/09/2020 18:54:21",0,"Cleanup from mbox formatting ""A lot of code from EVG-6592 cases on the fact that old commit queue items won't be mbox formatted. Can now go back and remove that.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9220","01/10/2020 14:43:34",1,"Manage Okta user cache ""Once the has functions to manage the cache, it should be possible to save access/refresh tokens for later use. The access token can be used to reauthorize the user periodically (like what is currently done for LDAP). Access tokens might be less useful to store since they are very short-lived, while refresh tokens have a much longer lifetime.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9224","01/10/2020 15:10:22",2,"Instrument Okta calls ""It would be useful to know how long re-authorization of a user takes after they've already logged in. Since users are periodically re-authorized, it would be good to know the impact on page load times. It would also be valuable to know how long each individual request takes.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7190","01/10/2020 18:15:59",0,"Commit queue merges without community changes get the wrong commit message ""If a commit queue patch is created that contains no community changes the message will be set to the default patch message [here|https://github.com/evergreen-ci/evergreen/blob/6fd48e04af41e4c05285cac628e4e67d8be95685/operations/patch_util.go#L100], instead of waiting for the message to get filled in when the modules are added. ""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7191","01/10/2020 19:19:59",1,"Stranded task job should be able to handle docker containers ""could not check host with timed out task for external termination: error getting cloud status for host container-1691131538922827018: Failed to get container information for host 'container-1691131538922827018': Docker inspect API call failed for container 'container-1691131538922827018': Cannot connect to the Docker daemon at tcp://ec2-34-205-72-241.compute-1.amazonaws.com:2376. Is the docker daemon running?""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7196","01/13/2020 02:47:03",2,"Determine caching mechanism for REST API responses ""It is likely that we will need to make REST API calls in some parts of our application. We should determine how we want to cache responses from these endpoints.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7205","01/13/2020 14:01:00",0,"Spruce Auth ""Protected routes redirect user to login page if not authenticated # Login page (nothing special since need designs for final version) # Auth context/state that holds global auth state about current user # Auth wrapper for protected routes # Query response middleware/afterware that updates auth context based on request response""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7199","01/13/2020 14:37:37",2,"Query for Evergreen projects ""Get all Evergreen projects to be displayed in the projects dropdown. Only project name and ID are needed.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7207","01/13/2020 14:38:28",2,"Query for user's favorites ""Query for user's favorites projects. Only project name and ID are needed.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7203","01/13/2020 14:48:50",2,"Query for individual task ""Given a task id get all data needed for that task Separate queries for task metadata and task tests, execution tasks, files and build baron since these will be fetched separately""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7209","01/13/2020 14:51:42",2,"Task page route ""Route for task page `/task/\{taskId}`""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7208","01/13/2020 14:56:09",0,"Tests table "".Queries for the tests for a task. This ticket only includes the table itself. It does NOT include the tabs, i.e. Tests, Execution Tasks, etc Shows a row for each test with Name, Status and Time. Showing logs for each test done in a separate ticket. Infinite scroll is also in another ticket. !Task Page - Tests - Branch.png!""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7201","01/13/2020 15:00:10",2,"Sorting tests table updates the URL query params ""When table is sorted, update url query params and re-fetch task tests data""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7206","01/13/2020 15:04:49",2,"Query for task tests ""Given a task ID and sort/filter params, fetch paginated task tests""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7210","01/13/2020 15:20:46",2,"Logs column in Tests table ""HTML button directs user to Lobster RAW button directs user to raw logs""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8631","01/13/2020 16:35:54",2,"add default reporting jobs to barque ""amboy job status cron setup and addr system stats ""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7213","01/13/2020 17:19:43",2,"Investigate large numbers of TerminateInstance calls to AWS ""We make many calls to TerminateInstances, even more than calls to CreateFleet and RunInstances combined. We should investigate if this is necessary since it eats up AWS API calls and increases host-intent queue length. The attached image is a search over the past week.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7214","01/13/2020 18:40:33",2,"Enable Copying redacted variables between projects ""Can we add a way to copy REDACTED variables into a different EXISTING project (and leave then redacted) Potentially, also have a “dry-run” option to tell if a value will or will not change ""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7215","01/13/2020 20:37:47",1,"Unable to add variables to a new project ""After creating a new project I tried adding two variables and saved the project. I get a message that the project was saved successfully, but the variables are not there. If I close down the browser and go back into the project in a new browser window, the new variables still do not show up.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9240","01/14/2020 14:22:33",2,"Add Okta mock tests ""In order to do this, we have to create a mock HTTP server that can return mock responses for each endpoint.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9153","01/14/2020 14:31:25",2,"Validate access and ID tokens, acquire refresh token ""The token can be easily verified by using the /introspect endpoint to validate the token. However, this requires one network call per authorization, which has to be done often. If we don't care about performance and number of requests made, this might be good enough. The best way to do this would be to add application logic for caching and periodic refreshing of the keys from the authorization server. If the access token is expired, get a new one using the refresh token. If the access token key does not match the authorization server key, try getting a new access token first. If the keys still do not match, only then do we try getting the authorization server keys. That way, will only ask for new keys if the validation fails (due to key mismatch) or if the caching TTL has expired.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7219","01/14/2020 16:20:19",2,"Increase time in queue factor & (maybe) reduce boost for mainline commits ""boost from 48 hrs to 12 or 24""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7220","01/15/2020 14:29:03",2,"Docker host terminated while running a task ""https://evergreen.mongodb.com/task/mongodb_mongo_master_ubuntu1604_container_cwrwc_rc_majority_passthrough_gen_patch_e32c7e7b2a4f8680090616c256b4cd8ad2a047ee_5e1c98e67742ae47a535cea1_20_01_13_16_20_55##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%2522e32c7e7b2a4f8680090616c256b4cd8ad2a047ee%2522%257D%255D%257D""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9198","01/15/2020 16:13:35",1,"Migrate poplar off of github.com/satori/go.uuid ""It has an insecure randomness vulnerability.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8701","01/16/2020 18:49:42",1,"Trigger signal processing job in cedar when YCSB added manually ""As a DAG engineer, I want YCSB metrics being added to cedar to trigger signal processing jobs. Such that, Signals get processed. AC: * YCSB signals processed when they are updated""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7226","01/17/2020 00:14:40",2,"s3.put should handle absolute paths to ""local_file"" ""The s3.put command seems to always concatenate the working directory to the {{local_file}} argument. It may be desirable to use the full path instead of a relative path because the path that {{local_file}} is relative to is not immediately obvious. See example output here where local_file is ${workdir}/mongod.log """," [2020/01/16 21:14:11.261] Running command 's3.put' (step 2 of 5) [2020/01/16 21:14:11.308] Putting /data/mci/ec0b598fee99990d3f7316852c57a61e/mongod.log into https://s3.amazonaws.com//mciuploads/perf/linux-wt-standalone/d6dad3428ca5773f8fcb7e75767e4817f7a185dd/performance_linux_wt_standalone_genny_scale_InsertRemove_patch_d6dad3428ca5773f8fcb7e75767e4817f7a185dd_5e20b96857e85a0902b35424_20_01_16_19_28_57/5e20b96857e85a0902b35424/logs/mongod-performance_linux_wt_standalone_patch_d6dad3428ca5773f8fcb7e75767e4817f7a185dd_5e20b96857e85a0902b35424_20_01_16_19_28_57.log [2020/01/16 21:14:11.308] performing s3 put to mciuploads of perf/linux-wt-standalone/d6dad3428ca5773f8fcb7e75767e4817f7a185dd/performance_linux_wt_standalone_genny_scale_InsertRemove_patch_d6dad3428ca5773f8fcb7e75767e4817f7a185dd_5e20b96857e85a0902b35424_20_01_16_19_28_57/5e20b96857e85a0902b35424/logs/mongod-performance_linux_wt_standalone_patch_d6dad3428ca5773f8fcb7e75767e4817f7a185dd_5e20b96857e85a0902b35424_20_01_16_19_28_57.log [1 of 10] [2020/01/16 21:14:11.308] Command failed: missing file /data/mci/ec0b598fee99990d3f7316852c57a61e/data/mci/ec0b598fee99990d3f7316852c57a61e/mongod.log: problem opening file /data/mci/ec0b598fee99990d3f7316852c57a61e/data/mci/ec0b598fee99990d3f7316852c57a61e/mongod.log: open /data/mci/ec0b598fee99990d3f7316852c57a61e/data/mci/ec0b598fee99990d3f7316852c57a61e/mongod.log: no such file or directory [2020/01/16 21:14:11.308] ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7227","01/17/2020 15:02:14",2,"optimize task gql query when testResults is omitted ""only query db for TestResults when testResults is included as a task query field include integration tests""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8348","01/17/2020 16:39:26",2,"GET Task mappings should return just display tasks, not execution tasks, for generated tasks that are split into display and execution tasks ""Background: The GET /projects/\{project\}/task-mappings endpoint returns *all* tasks that are related to a list of changed source files. This means that if there is a display task """"replica_sets_large_txns_format"""" that contains numerous executions tasks (i.e. """"replica_sets_large_txns_format_01_enterprise-rhel-62-64-bit"""", """"replica_sets_large_txns_format_02_enterprise-rhel-62-64-bit""""), both the display task and the execution tasks are returned by the endpoint. You can see an example of this if you make a GET /projects/\{project\}/task-mappings request with the src/mongo/db/storage/storage_engine_impl.cpp changed file. User story: As a MongoDB engineer, I should be confident that, for generated tasks that are split into display and execution tasks, the selected-tests GET /projects/\{project\}/task-mappings endpoint only returns the display task, and not a display tasks's execution tasks, so that I am not being returned duplicate tasks to run. AC: * Making a GET /projects/\{project\}/task-mappings request with the src/mongo/db/storage/storage_engine_impl.cpp changed file should return the """"replica_sets_large_txns_format"""" task but not its execution tasks""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7229","01/18/2020 15:10:11",2,"BFGs are created for every stepback commit ""It look like BFGs are being created for every stepback commit instead of the first one that failed. One example is BFG-393900 and BFG-394834. They were created from two consecutive commits on the same variant with the same failing task name and test name. I have not checked other occurrences of stepback since Jira search in the build baron plugin has been timing out for me. This is causing pain for build barons and it would be great if it could be prioritized.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7230","01/20/2020 23:46:46",2,"populate logs field in Task gql query ""populate the logs field in Task gql query in Task resolver""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7235","01/21/2020 14:53:53",0,"Consider region when creating a client ""Now modify GetManagerOptions and fromDistroSettings to choose ProviderSettings by region, although initially can set this always to a default, and potentially other places.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7234","01/21/2020 14:53:53",0,"Cleanup from ProviderSettings move ""Potentially two fields will be considered while moving data in provider settings to a new provider settings variable, so after some point we need to go back and remove support for the old field.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7238","01/21/2020 14:53:54",1,"Add region option to spawn host API routes ""Modify spawn host routes in service.go""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7237","01/21/2020 14:53:54",1,"Specify default spawn host region in User Settings ""Should default to us-east-1, our current default""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7236","01/21/2020 14:53:54",2,"Consider region for Hosts ""Add Region to Host struct, and consider this region when creating a client for a given host document (to create/modify/query a host)""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7241","01/21/2020 14:53:55",1,"Add flag to filter CLI `host list` by region ""Should be optional""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7240","01/21/2020 14:53:55",2,"Add region option to spawn host page ""Should default to the region in UserSettings.If this region isn’t available for this instance type, should return a helpful error to the user.Add region (or perhaps more helpfully availability zone) to the Spawn Host details card.Should also modify the UI API routes.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7239","01/21/2020 14:53:55",1,"Add region option to spawn host CLI ""Should default to the region in UserSettings.If this region isn’t available for this instance type, should return a helpful error to the user.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7242","01/21/2020 18:20:33",2,"Improve timeouts in Evergreen ""Timeouts can be confusing, and it is not obvious to users which one was hit.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7247","01/21/2020 20:59:56",2,"Require explicit confirmation for unexpirable instance and volume deletion ""Perhaps have users type """"delete"""" if they really want to delete, i.e. less automatic than typing y to our usual prompts. This should handle the spawn host page as well as the CLI.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7248","01/21/2020 21:18:33",1,"Cleanup the string copy of a patch from the upload ""When we've concluded most users have updated their CLI past [f03612|https://github.com/evergreen-ci/evergreen/commit/f036124c01bd7de98e2b03e660636d08634a6fe9] we should remove the extra copy of the patch from the upload.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7249","01/21/2020 22:35:47",2,"Docker containers get decommissioned too aggressively ""https://evergreen.mongodb.com/host/container-3894847770651935761""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7250","01/21/2020 22:43:34",1,"Validator should check for empty functions ""Functions without commands cause panics bc we assume that they don't exist (but also they shouldn't exist so the validator should handle this)""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8710","01/23/2020 21:15:50",1,"Add index for signal processing to cedar ""As a DAG engineer, I'd like an index to cover the queries I issue during signal processing. AC: * Cover get performance data queries with an appropriate index""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8722","01/24/2020 14:52:15",1,"Null coalesce rollup statistic values to 0 ""We have experienced failure processing signals in cedar when getting time series with null values in them --> Golang bson driver does not know how to convert these to floats. As a DAG engineer, I'd like to coalesce these nulls to 0, such that I can process signals! AC: * Null coalesce to statistic values to 0 during performance data aggregation.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8697","01/24/2020 15:45:06",1,"Log url and auth user when signal processing client fails to detect change points ""As a DAG engineer, I'd like to log out service url and auth user when failing to detect change points. Such that, I can debug failures.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7262","01/24/2020 16:58:39",2,"Commit queue tasks aborted by the failure of another task should indicate what task had a real failure ""as it is people waste a lot of time trying to track down the actual error message.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7263","01/24/2020 17:16:51",2,"Automatically authenticate user ""Right now, user must login to be authenticated even if they have an evergreen cookie in their browser. The app should automatically authenticate the user if a cookie is present and the user can make successful requests to the API. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7264","01/24/2020 18:19:24",0,"configurable logging to s3 should not use environment variables ""either shared credentials file or IAM roles.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8693","01/24/2020 18:21:26",1,"Add resource limits to selected-tests ""We should implement resource limits in Kanopy as described here: https://github.com/10gen/kanopy-docs/blob/master/docs/production_checklist.md#resource-requestslimits See this ticket for more information about default resource limits: https://jira.mongodb.org/browse/TIG-2051 User story: As a DAG engineer, I should know that there are resource limits set on the selected-tests service, so that I am following the process advised by the Kanopy docs and so that resources and resource requests are distributed properly across services. AC: * selected-tests has resource limits for cpu and memory * resource limits and resource requests are distributed across services""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8577","01/24/2020 18:24:33",2,"Expose analysis results through cedar API ""As a DAG engineer, I'd like to expose analysis results through cedar's API, such that, consumers can see them! AC: * expose analysis results through cedar's API""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7271","01/24/2020 20:11:32",2,"Schedule task mutation ""Mutation to schedule a task""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7274","01/24/2020 20:12:43",2,"Restart task mutation ""mutation to restart task""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8835","01/27/2020 13:55:30",2,"Try using library for access token validator instead of endpoint ""Access tokens can be validated by polling the introspect endpoint but it's slower than using the Okta library to verify the access token. For some reason, the access token doesn't validate when used with the library (it says that its {{kid}}, which is a value used to check which public key to use for decoding, does not match any of the keys returned by the authorization server).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8739","01/27/2020 14:46:24",0,"Add rate limited jobs to Amboy ""Right now, there exists a concept of rate limited jobs in the sense that you can set their id with a rounded off timestamp, generating a duplicate key error on collision that you then ignore. It would be nice if we had some more formal notion of a rate limited job, that doesn't generate errors to handle in the specific case that it is enqueued when another of its type is still in the queue.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8555","01/27/2020 14:48:01",1,"Don't generate errors due to duplicate keys for recalculate-change-points job queueing ""As a DAG engineer, I'd like not to generate errors when failing to enqueue a job due to rate limiting. Such that, I don't get errors about a system that functions properly! AC: * No errors generated due to rate limiting""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7280","01/27/2020 15:28:16",2,"Query for a Task's files ""query for a tasks files. the files should be grouped by exec task""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7433","01/27/2020 15:34:02",2,"Allow downloading of very long log files ""Using curl (with no timeout)""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7284","01/27/2020 19:42:22",0,"Fix copied host allocator settings ""It has a misspelling in the JS file so it isn't be copied when we clone the distro.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7288","01/28/2020 15:45:50",1,"Creating a new project doesn't make it visible in the UI immediately ""Also it adds it to the list of all projects twice""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7289","01/28/2020 16:28:32",2,"Local evergreen server is unresponsive in CI environment ""Making calls to local evergreen server in CI results in a 500 error and an ECONREFUSED error because the local evergreen server is unresponsive. Figure out how to successfully talk to local evg server in ci. ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8342","01/28/2020 16:34:45",0,"LDAP URL is ambiguous ""Since URL parsing is ambiguous, checking the scheme in the LDAP url is insufficient since it might interpret host:port format as the scheme.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8611","01/28/2020 16:40:44",1,"Fix periodic change point job ""As a DAG engineer, I'd like to fix the periodic change point job, such that, it works. AC: * Periodic change point job works""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8390","01/28/2020 16:49:28",0,"Specify cookie domain when clearing it ""If the cookie domain is not specified when cookies are cleared, it will fail to clear the correct cookie because it will assume the domain is the current domain.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8708","01/28/2020 17:23:43",1,"Handle case of no measurements ""As a DAG engineer, I want signal processing for performance results with no recorded measurements not to panic, such that, people don't have to deal with meaningless panics. AC: * GetPerformanceData should return empty performance results if no measurements are recorded in the DB.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7290","01/28/2020 18:35:32",1,"Catch nil when no user is found ""[Here|https://github.com/evergreen-ci/evergreen/blob/6e3a626a35b044c27bf3af4c62b244d307452a4f/model/project_ref.go#L718]""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8772","01/28/2020 19:10:41",1,"Support sending SIGABRT in Jasper grpc ""Resmoke sends SIGABRT to processes that are configured to archive in the event of a test failure in order to force the creation of a core dump. This doesn't currently seem to be [supported|https://github.com/mongodb/jasper/blob/76466aca6b613c14629889d49c53a5e41efd0655/jasper.proto#L133] for Jasper processes. ""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7292","01/28/2020 20:54:09",2,"Under load notifications can take a long time to be processed ""We rate limit ourselves to 1000 notifications per notification run, which seems to be not enough. We should be cautious scaling these up though, since this is natural rate limiting of queries against the DB.""","",0,0,0,0,0,0,1,0,0,1,0,0,0,0,0
+"EVG-8658","01/30/2020 14:52:57",1,"Retry 499 status codes in change detector ""Sometimes the load balancer for Kanopy returns 499 for timeouts. We should retry these requests since it is usually due to burst activity.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7304","01/30/2020 23:40:14",1,"File one BFG per Display Task Instead of Execution Task ""With the Server test suites moving to execution tasks for increased parallellism, the number of BFG tickets has grown accordingly. This is caused by a couple of factors. First, widespread failures that affect multiple tests in a suite would create X number of BFGs instead of 1. Second, consistently failing tests are not guaranteed to run in the same execution task, due to the way the knapsack packing algorithm relies on test history information. This prevents backtracking from working and could cause more BFGs to be created for the same failure on different execution tasks""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9031","01/31/2020 18:55:46",1,"Remove group check from Okta ""It seems like any users who can log into Okta is an employee (which is all we care about), so we don't have to perform the group check that is done in LDAP Therefore, we can remove (or make optional) much of the logic with checking the user's groups from userinfo and validating the access token. It seems like this is somewhat annoying to do because the ID token validation expects a nonce to be provided, but there is none for an ID token received from refreshing tokens. There is [some ambiguity|https://github.com/IdentityServer/IdentityServer4/issues/2180] about what to do since the ID token doesn't have a nonce when we refresh tokens. [This thread|https://bitbucket.org/openid/connect/issues/1025/ambiguity-with-how-nonce-is-handled-on#comment-44979149] suggests that a refresh grant should not return a nonce, so the Okta library may simply be wrong.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8619","01/31/2020 21:49:45",1,"Reconfigure change point detector retry policy to be lenient ""The change point detector currently retries requests 10 times with a maximum delay between requests of 5 seconds. This is not allowing signal processing service to recover in time when a 499 comes up. As a DAG engineer, I'd like to reconfigure the detector to retry 50 times, with a max delay of 5 minutes, Such that, the service has time to recover and respond during high load periods. AC: * change point detector reconfigured per above.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7309","01/31/2020 22:36:07",0,"Mutation to remove favorite project from user ""Given a project id, remove project from user's favorites""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7311","02/03/2020 14:13:49",1,"Only import required antd css ""Install react-app-rewired customize-cra and babel-plugin-import for selective importing [https://ant.design/docs/react/use-with-create-react-app#Advanced-Guides]""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7313","02/03/2020 16:00:22",2,"TaskTests query returns duplicate keys between paginated queries ""TaskTests query returns duplicate keys across two subsequent pages with the same sort category, direction, and limit ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7314","02/03/2020 21:47:04",1,"Failure: test-model-grid on Ubuntu 16.04 (Docker) [Evergreen @ 924ea8ee] (TestFetchFailures) "" h2. [test-model-grid failed on Ubuntu 16.04 (Docker)|https://evergreen.mongodb.com/task/evergreen_ubuntu1604_docker_test_model_grid_924ea8ee4016255a2cdf7496ab5b8284f1fbb899_20_01_06_17_05_58/0] Host: [|https://evergreen.mongodb.com/host/container-34966100467904843] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/924ea8ee4016255a2cdf7496ab5b8284f1fbb899]: Revert """"Revert """"Revert """"Revert """"EVG-7119: fix toolchain issues (#2998)"""""""""""""""" This reverts commit 1bfa579f24aff033c49866945ec3f8e71e92f90d. | 06 Jan 20 17:05 UTC Evergreen Subscription: ; Evergreen Event: *TestFetchFailures* - [Logs|https://evergreen.mongodb.com/test_log/5e138898850e611b0b739ca4] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_docker_test_model_grid_924ea8ee4016255a2cdf7496ab5b8284f1fbb899_20_01_06_17_05_58#TestFetchFailures=fail] [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_ubuntu1604_docker_test_model_grid_924ea8ee4016255a2cdf7496ab5b8284f1fbb899_20_01_06_17_05_58/0?type=T ] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-7316","02/05/2020 15:04:05",2,"Add UTC as a timezone ""Please add UTC as a timezone in the Evergreen UI so that my Evergreen UI and my logs from build slaves are in the same timezone.""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8485","02/05/2020 21:30:15",2,"Daily update script is creating duplicate mappings instead of updating pre-existing mappings ""Currently, the cronjobs that run daily to update the selected-tests task mappings and test mappings have a bug. When the past day's evergreen versions or git commits, respectively, are analyzed, the newly created mappings are inserted into the db like so: https://github.com/mongodb/selected-tests/blob/a8251a6ab0f880526d7ac27f545c389f3a87c964/src/selectedtests/task_mappings/update_task_mappings.py#L41. This causes an issue because, for source_files with pre-existing mappings, the mapping should be updated, not inserted. The resulting duplicates are visible from running these queries: User story: As a MongoDB engineer, I should know that there will not be duplicate test mappings or task mappings in the selected-tests database, so that the data which I am using to run my selected tests patch builds is clean. AC: * No one source_file should have two entries in either the test_mappings or task_mappings collection * Put unique index on source file field * Clean database and rerun code once bug is fixed"""," db.test_mappings.aggregate({ $group: {_id: {source_file: '$source_file'}, count: {$sum: 1}}}) db.task_mappings.aggregate({ $group: {_id: {source_file: '$source_file'}, count: {$sum: 1}}}) ",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7318","02/05/2020 23:33:05",2,"Figure out how to get Cypress to match on the body of a Post ""There are probably ways we can workaround this issue [https://github.com/cypress-io/cypress/issues/3755]""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7320","02/06/2020 09:24:35",1,"Difficulties changing timezone ""I wanted to change timezone to """"Central European Time"""", but I couldn't due to false-negative form validation (error says """"Please fill out all required fields before submitting""""): !timezone.png|thumbnail! I turns out the fields requiring values are hidden and only visible if I toggle on and off the """"Opt In to New Patch Page"""" checkbox. !required-fields.png|thumbnail! It might be related that I previously toggled this checkbox on to see how the new patch page looked.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,1,0
+"EVG-7321","02/06/2020 13:51:06",2,"Running command 'archive.targz_extract' fails on the Node.js archive ""I am downloading Node.js in a specific version: And I was trying to extract it using the “archive.targz_extract” command in evergreen: But I get the following error: The error seem to be thrown if an unexpected type of item is being extracted from the archive (see [https://github.com/evergreen-ci/evergreen/blob/master/util/artifacts_tar.go#L209-L211]). Are the any (undocumented) limitations to the “archive.targz_extract” command?""","curl https://nodejs.org/dist/v10.19.0/node-v10.19.0-darwin-x64.tar.gz --output node.tgz - command: archive.targz_extract params: path: """"node.tgz"""" destination: """"node""""[2020/02/06 13:11:05.375] Running command 'archive.targz_extract' in """"install node.js"""" (step 2.2 of 5) [2020/02/06 13:11:06.196] Command failed: problem extracting '/data/mci/07462cdf3e08015e1aea828c8ca41aa5/node.tgz': error extracting /data/mci/07462cdf3e08015e1aea828c8ca41aa5/node: Unknown file type in archive. [2020/02/06 13:11:06.196] Task completed - FAILURE.",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9204","02/06/2020 16:26:34",2,"migrate cedar and barque to simple service users ""Ideally, we would use Okta service users for this. However, since Okta service accounts are not ready yet and LDAP is imminently going to lose support, we have to implement service users in the same way as is currently done for evergreen as a stopgap. Note: after the mgiration is complete, regular users who were relying on their LDAP credentials will no longer be able to auth and they'll have to ask us for service users.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7323","02/06/2020 16:59:12",1,"Commit queue includes file summary in commit messages ""If a commit contains new or moved files the commit message includes a record of the added/moved files.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7327","02/07/2020 02:13:44",2,"Query to retrieve a single patch ""Given a patch id, get the patch metadata""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7328","02/07/2020 02:15:25",2,"Add version displayName to task query ""The breadcrumb on the task route needs to show the displayName of the patch it belongs to. Add the version's displayName to the task query so it can be passed to the TaskBreadcrumb component""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7329","02/07/2020 14:53:11",0,"Distro API routes should use multiple provider settings ""When creating/modifying distros. Should also check that the default instance type is valid for the region.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7331","02/07/2020 18:40:21",2,"Migrate usernames and deploy Okta to prod ""Once we turn on Okta, we have to migrate all the non-conformant LDAP names to their new Okta names.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7332","02/07/2020 18:42:03",1,"Enable mac static host provisioning ""Should start on the distros with the smallest host pools to increase confidence.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7333","02/07/2020 18:43:52",2,"Write docs for virtual workstations ""This goes in the employee wiki.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7334","02/07/2020 18:46:25",2,"Clean up host provisioning ""For example, some of the logic done in the provisioning setup job should probably be cleaned up.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7335","02/07/2020 21:25:55",0,"Make static host authorized keys path configurable ""Right now it's hard-coded as {{~/.ssh/authorized_keys2}}, but this can be changed.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7336","02/07/2020 21:46:26",1,"Re-add vet shadow to linters ""It's currently disabled. The previous incarnation of it in gometalinter was enabled, but half-baked and barely worked.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7337","02/08/2020 00:02:06",1,"Toggle from new UI to legacy UI ""* If I am on 'my-patches' route then clicking toggle should bring me to 'my patches' in legacy UI * If I am on 'patch' route then clicking toggle should bring me to 'patch' in legacy UI * If I am on 'task' route then clicking toggle should bring me to 'task' in legacy UI""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7338","02/08/2020 00:03:56",1,"Toggle from legacy UI to new UI ""PAGES UNSUPPORTED BY SPRUCE WILL NOT HAVE A LINK IN THE NAVBAR Link in navbar that goes to corresponding Spruce page""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7339","02/08/2020 00:09:16",2,"Layout of the Task and Patch routes ""The patch and task routes have the same layout. # Header # Left sider # Content Use antd layout""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7340","02/08/2020 00:10:39",2,"Catch-all route that shows 404 for unknown routes ""If I go to a route that does not exist then I should see a 404 page ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7341","02/08/2020 00:15:00",2,"Toast feature to show errors ""When an occurs, such as an internal server error from trying to favorite a project, show a toast message for that error. Use context to implement. Should be able to use it like: """," try { doSomethingAsync() } catch(e) { toast.error(e.message) }",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7343","02/08/2020 01:24:16",0,"Patch header and metadata ""Use data from patch query to render the patch name in header and the metadata in left sider""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7344","02/08/2020 01:25:17",2,"Query for patch tasks ""given a patch id, get all tasks for that patch in the shape specific in UI for tasks table""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7345","02/08/2020 01:32:50",2,"Patch tasks table ""Use data from patch tasks query to render the tasks table.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7346","02/08/2020 01:36:11",2,"Query for patch tasks sorted by name ""Accept a name parameter in query for name that is either ascending or descending.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7347","02/08/2020 01:36:54",2,"Query for patch tasks sorted by status ""Accept param in query to sort by ascending or descending""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7348","02/08/2020 01:37:23",2,"Query for patch tasks sorted by variant ""Accept param in query to sort by variant asc or desc""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7351","02/08/2020 01:40:47",1,"Status of patch auto updates ""When the status of a patch updates, the UI should update to reflect it""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7353","02/08/2020 01:45:41",2,"Add modify version mutations to patch buttons ""Show buttons for schedule, unschedule, abort and restarting a patch. Disable buttons based on status. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7359","02/10/2020 13:39:01",2,"Upgrade to Cypress 4.0 ""Cypress is currently at 3.8.2 which has known issues running on Ubuntu 18.04. Upgrade Cypress to version 4.0 and then change the variant in Evergreen from Ubuntu 16.04 to 18.04. Cypress 4.0 fixes known bug in Ubuntu 1604.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-9267","02/10/2020 14:18:01",2,"Use golangci-lint in all repos ""This is already being used in the main evergreen repo and in Jasper.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7360","02/10/2020 15:05:30",0,"Preserve reprovisioning state until host is reprovisioned ""Right now, the host allocator will clear the reprovisioning state without checking if the host already needs reprovisioning.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8917","02/10/2020 16:43:23",1,"Fix windows file handling in tests ""Test failure: https://evergreen.mongodb.com/task/jasper_windows_test_jasper_1475c6ddedb0af97b9a57b1bed704da25ac5f3cc_20_02_09_19_20_08""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7362","02/10/2020 16:55:00",1,"Create alert for 30% of GitHub API limit ""We currently don't have good visibility into how much of our GitHub API limit we're using. As more projects enter Evergreen, we may get close to our limit and not realize it. We should make the log statement structured. (Currently it looks like """"message: Github API response: 200 OK. Rate limit: 4412/5000"""".) And we should create an alert for if we get below some threshold (1500?). Once that happens, we should begin planning our approach.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7363","02/10/2020 16:55:47",1,"Full page loading screen ""Currently the full page loading screen is just the work """"LOADING..."""". We should get designs to show a loading spinner.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8640","02/10/2020 19:22:51",2,"Fix timeout handling in cedar change point detector ""Sometimes our service times out and returns a 499 status code. This appears to not make it through to cedar, and an [error is generated|https://github.com/evergreen-ci/cedar/blob/master/perf/change_point_detector.go#L136]. Probably the fix consists of configuring the client timeout properly. As a DAG engineer, I'd like to fix whatever is going wrong, such that, it works. AC: * Retrying timeouts properly""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7365","02/11/2020 13:33:04",1,"Fix rsync test on Windows ""The rsync binary on Windows is the cygwin rsync binary, so it only accepts cygwin paths and not native Windows paths.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7366","02/11/2020 14:42:49",1,"Setting a host to unexpirable panics """""," gimlet/middleware_grip.go:177 runtime/asm_amd64.s:523 runtime/panic.go:513 runtime/panic.go:82 runtime/signal_unix.go:390 service/spawn.go:318 service/ui.go:389 net/http/server.go:1964 negroni/negroni.go:46 negroni/negroni.go:29 negroni/negroni.go:38 negroni/negroni.go:38 service/middleware.go:260 gimlet/middleware_wrapper.go:16 negroni/negroni.go:38 negroni/negroni.go:38 service/middleware.go:136 gimlet/middleware_wrapper.go:16 negroni/negroni.go:38 negroni/negroni.go:86 mux/mux.go:212 negroni/negroni.go:46 negroni/negroni.go:29 negroni/negroni.go:38 negroni/negroni.go:38 negroni/static.go:34 negroni/negroni.go:38 negroni/negroni.go:38 negroni/static.go:34 negroni/negroni.go:38 negroni/negroni.go:38 gimlet/middleware_auth.go:37 negroni/negroni.go:38 negroni/negroni.go:38 gimlet/middleware_auth_user.go:209 negroni/negroni.go:38 negroni/negroni.go:38 gimlet/middleware_grip.go:192 negroni/negroni.go:38 negroni/negroni.go:86 net/http/server.go:2741 net/http/server.go:1847 runtime/asm_amd64.s:1333 ",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7371","02/11/2020 18:59:11",1,"Commit Queue strips windows line endings from patches ""If a patch is submitted with a windows line ending the extra \r is removed from the patch by git mailsplit. This causes 1) patches to not apply, and 2) lines to be committed without windows line endings.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8782","02/11/2020 19:53:54",1,"Use new putunique API in cedar/amboy ""As a DAG, I would like to use the new putunique API from amboy in cedar, such that, we don't generate errors due to duplicate jobs being enqueued (these trigger alerts). AC: * Use new putunique API when enqueueing recalculate change points jobs.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7376","02/12/2020 11:16:20",2,"Enterprise module fails to compile - manifest lost? ""Consider following 3 builds: Base commit, Jan 31: https://evergreen.mongodb.com/version/sys_perf_23603867ca8e968453683a9dc55c72a32e3cb0b5 Patch 1, Jan 31: https://evergreen.mongodb.com/version/5e3435cbe3c33137bd061b9c Patch 2, Feb 10: https://evergreen.mongodb.com/version/5e416147e3c331076ef691b7 Manifest shows following revisions for enterprise module: Base: 4ef3662babc Patch 1: 4ef3662babc Patch 2: af509c51bc2373 Consequently, Patch 2 fails to compile.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7377","02/12/2020 15:20:16",1,"Check if agent monitor is still alive before attempting deploy ""Most agent monitor deploys that occur during app server deploys are not actually necessary, especially during high times of churn (i.e. daily app server deploys). The job should no-op if the agent monitor is still running in Jasper, and assume the agent will come up and contact the app server if given enough time.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8615","02/12/2020 15:24:04",1,"Handle inconsistent timestamp format from the evergreen api ""The evergreen API returns different timestamp format for the """"create_time"""" field. It sometimes includes milliseconds and sometimes does not. Which means that we frequently fail to parse the timestamp because we are getting an expected format. We need to be able to detect when the format is different and handle it appropriately. ---- As an evergreen api user I want the api client to be able to handle different date formats So that I can handle when evergreen returns different date formats. ---- AC: * The api client can handle when evergreen returns date format without milliseconds as well as when it returns them with milliseconds.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7378","02/12/2020 16:24:32",1,"Failure: coverage on Coverage [Evergreen @ 7ec94b3e] (TestGetRevisionsSinceWithPaging) ""h2. [coverage failed on Coverage|https://evergreen.mongodb.com/task/evergreen_coverage_coverage_7ec94b3eb3377a60aa93fffc9cf8feae8261670f_20_02_12_14_14_50/0] Host: [ec2-34-207-144-162.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-0e949e59eb0cca13c] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/7ec94b3eb3377a60aa93fffc9cf8feae8261670f]: Increase time until unresponsive agent monitor attempts to deploy (#3149) | 12 Feb 20 14:14 UTC Evergreen Subscription: ; Evergreen Event: *TestGetRevisionsSinceWithPaging* - [Logs|https://evergreen.mongodb.com/test_log/5e440ac77742ae6aeb621482] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_coverage_coverage_7ec94b3eb3377a60aa93fffc9cf8feae8261670f_20_02_12_14_14_50#TestGetRevisionsSinceWithPaging=fail] [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_coverage_coverage_7ec94b3eb3377a60aa93fffc9cf8feae8261670f_20_02_12_14_14_50/0?type=T ] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-7379","02/12/2020 16:39:37",2,"Failure: test-rest-client on Windows [Evergreen @ 7ec94b3e] (TestLogSenders) ""h2. [test-rest-client failed on Windows|https://evergreen.mongodb.com/task/evergreen_windows_test_rest_client_7ec94b3eb3377a60aa93fffc9cf8feae8261670f_20_02_12_14_14_50/0] Host: [ec2-54-242-24-13.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-02e538b76ca87fa7b] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/7ec94b3eb3377a60aa93fffc9cf8feae8261670f]: Increase time until unresponsive agent monitor attempts to deploy (#3149) | 12 Feb 20 14:14 UTC Evergreen Subscription: ; Evergreen Event: *TestLogSenders* - [Logs|https://evergreen.mongodb.com/test_log/5e44093b306615219ac16c40] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_rest_client_7ec94b3eb3377a60aa93fffc9cf8feae8261670f_20_02_12_14_14_50#TestLogSenders=fail] [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_windows_test_rest_client_7ec94b3eb3377a60aa93fffc9cf8feae8261670f_20_02_12_14_14_50/0?type=T ] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-7380","02/12/2020 19:32:27",1,"Reusable typography components ""Each piece of text in the app should be able to be classified. For example, """"My Patches"""" on patches page, and display name on task page are both h1's. The task/test names in the table are the same. Instead of assigning font size to each typography element rendered, we should be able to render a typography element. # Classify each piece of typography # Make reusable typography components based on those classifications""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7385","02/13/2020 15:26:08",1,"Add Validation to attach_artifact ""If a user attaches an artifact through the attach_artifact command and supplies their own link to a file, as opposed to doing it through the s3_put command, they should not be able to set Visibility to """"signed"""". The motivation: When a user attaches a file that way, it can be a link to GitHub, a wiki page, or any arbitrary link. Since it's not a url that we generated and uploaded to s3, we cannot sign it. Even if the link they are provided IS an s3 link, we still can't sign it, unless we expand the params to also take in the aws key and secret that they used to upload the file with. However, in that case, if that is the functionality that they are looking for, it would make more sense for them to upload it using the s3_put command than to hack attach_artifacts to behave like s3_put with visibility set to signed. ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7387","02/13/2020 19:38:48",1,"Evergreen fetch for Windows spawn host uses incorrect path ""The working directory needs to be a native Windows path in order to open the file, but it's actually specified as a cygwin-style (i.e. Unix) directory path. This has not been a problem up until now because almost all files go into {{/data}}, which is fine because {{/data}} can be found at the root of the C drive as well. This only affects legacy SSH hosts.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7388","02/13/2020 21:42:38",0,"Query for task logs ""GQL Query for task logs in Task Page ""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7389","02/13/2020 21:45:11",2,"render task logs in task page ""Render logs under the logs tab like so:""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7390","02/13/2020 21:48:03",2,"Files table in tasks page ""create a table with filter support !Screen Shot 2020-02-13 at 4.46.56 PM.png!""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7393","02/13/2020 22:22:00",2,"Task metadata panel ""render task metadata side panel. confirm the displayed in the screen shot is from a finalized design ""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8944","02/14/2020 12:39:28",0,"Run ssh client with non-interactive flags ""Since the SSH client runs without user input, it has to disable pty allocation and logs.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7396","02/14/2020 16:29:52",0,"Move local SSH key update job population ""It should go in `startSystemCronJobs`.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7397","02/14/2020 21:34:06",1,"Version build errors are not exposed in API ""If I go to https://evergreen.mongodb.com/version/mongoid_7.0.0_53e15f580edc4f281b826327ecbaff8e10d8c9a6 I see this error message: This information appears to be written into the page html, as follows: When I retrieve this version via the API, it has no error information: Is there another place in API where this error information is exposed? If not can it be added to the API please?"""," 1 error in configuration file - 'mongoid-7.0.0' project's 'install dependencies' function contains no commands (byebug) @version.info {""""version_id""""=>""""mongoid_7.0.0_53e15f580edc4f281b826327ecbaff8e10d8c9a6"""", """"create_time""""=>""""2020-02-14T19:34:54Z"""", """"start_time""""=>nil, """"finish_time""""=>nil, """"revision""""=>""""53e15f580edc4f281b826327ecbaff8e10d8c9a6"""", """"order""""=>114, """"project""""=>""""mongoid-7.0.0"""", """"author""""=>""""Oleg Pudeyev"""", """"author_email""""=>""""39304720+p-mongo@users.noreply.github.com"""", """"message""""=>""""MONGOID-4833 omit testing on ruby-head for now (#4725)\n\nCo-authored-by: Oleg Pudeyev """", """"status""""=>""""created"""", """"repo""""=>""""mongoid"""", """"branch""""=>""""7.0.0-stable"""", """"build_variants_status""""=>nil, """"requester""""=>""""gitter_request""""} ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7400","02/18/2020 19:01:23",2,"Empty patches error in git.getproject ""The `git.getproject` command is trying to apply empty patches.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8361","02/19/2020 19:54:25",2,"Handle inconsistent timestamps and missing manifests in cronjobs/process_task_mapping_work_items.sh script ""I was running the cronjobs/process_task_mapping_work_items.sh scripts today and noticed some new errors. Evergreen has changed their create_time format (addressed in MAKE-1156) so we need to use a more updated version of evergreen.py. Additionally, create_time is now stored in a format that is UTC offset-aware. Finally, evergreen versions like [this one|[https://evergreen.mongodb.com/rest/v2/versions/mongodb_mongo_master_f31bc89f66632b2d521be2d076dc23f94ff663eb]] raise an [error when getting manifest|[https://evergreen.mongodb.com/plugin/manifest/get/mongodb-mongo-master/f31bc89f66632b2d521be2d076dc23f94ff663eb]], so we need to handle this exception better. As a mongoDB engineer, I should be able to add my project to the selected-tests task mappings queue, and know that it will process 6 months of data without erroring, so that I have task mappings to work with. AC: * Running the following steps does not error: """," curl -X POST """"http://localhost:8080/projects/mongodb-mongo-master/task-mappings"""" -H """"accept: application/json"""" -H """"Content-Type: application/json"""" -d """"{ \""""source_file_regex\"""": \""""^src/mongo\"""", \""""module\"""": \""""enterprise\"""", \""""module_source_file_regex\"""": \""""^src\"""", \""""build_variant_regex\"""": \""""^\!\""""}"""" ./cronjobs/process_task_mapping_work_items.sh ",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7408","02/19/2020 20:38:53",2,"Task is blocked on duplicate dependencies ""A task was created with a duplicated dependency. When the depended-on task failed the three copies of the dependency were all set unattainable in the depending task. When the depended-on task was restarted and passed only the first instance of the dependency was set unattainable false in the depending task, so it's still blocked and won't run. Arguably the main issue is the duplicated dependency.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7412","02/20/2020 14:48:56",1,"Set retry limit on background reauthorization ""If it fails after some number of attempts, it should log the user out (i.e. clear the login cache).""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7413","02/20/2020 19:01:21",2,"Filter test table based on test name input and status dropdown ""taskTests query already supports these things, these are Spruce changes only. should store these options as query params so we can link to the table state directly""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7415","02/20/2020 19:40:44",2,"GQL Query for patch code changes ""query for code changes by patch id ""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7422","02/21/2020 13:02:29",2,"Allow task group setup to be pulled from s3 ""Once the setup has run for one task group, it should be uploaded to s3 so other hosts can pull the setup without having to run through the entire setup.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7423","02/21/2020 13:05:18",2,"Add command to restore directory using pail ""Once the agent has uploaded the working directory to s3, we should be able to retrieve it by its task/execution.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7424","02/21/2020 13:10:01",2,"Upload task working directory to s3 ""The agent should use the command to save a directory using pail to upload the task's working directory.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8988","02/21/2020 13:12:00",2,"Run test patches using barque ""This will iron out any bugs in the service in preparation for using it for releases.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7426","02/21/2020 15:30:33",2,"Add icecream scheduler domain name to distro settings and add route to edit config file ""We have to do this so that icecream compile nodes can get a config with the correct domain name to connect to the load balancer.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8720","02/21/2020 16:54:33",0,"Create manager that allows multiple auth methods ""We should be able to combine auth methods into a single user manager. Multiple auth only applies to getting users, not creating new users (to avoid possibly creating multiple users).""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7428","02/21/2020 17:01:31",2,"Investigate slow task performance on macos with new provisioning ""Example task failure: https://evergreen.mongodb.com/task/mongo_c_driver_darwin_debug_compile_compression_zlib_0bcde05863e482aa08adf29a9a1b98fd533f8942_20_02_19_19_56_09""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-7429","02/21/2020 18:21:32",1,"Virtual workstations should be unexpirable by default ""If the user has unexpirable hosts remaining a virtual workstation should be set to unexpirable on creation.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9254","02/21/2020 18:48:14",1,"Make launch daemons with user security session ""[This StackOverflow question|https://stackoverflow.com/questions/47582989/launchd-not-able-to-access-mac-os-keychains] (also [this one|https://stackoverflow.com/questions/6827874/missing-certificates-and-keys-in-the-keychain-while-using-jenkins-hudson-as-cont/9482707#9482707]) says that we need to use {{SessionCreate}} in order to have access to the user login keychain (i.e. {{~/Library/Keychains/login.keychain-db}}).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7432","02/21/2020 19:56:19",2,"Enable CORS on evg auth endpoints ""The auth endpoints are the only REST endpoints that Spruce talks to (aside from the gql endpoint). These endpoints will need to allow CORS from spruce.mongodb.com to work in staging and in production. When I first tried working on this, I ran into issues with auth not working as expected anymore. I was able to login and logout, however the user cookie was not being removed on logout. This caused tests to fail.""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-9239","02/24/2020 14:24:56",0,"Convert usernames when making requests against LDAP ""We have to truncate usernames to be firstname.lastname for compatibility with the existing evergreen users. As the inverse operation, we have to also reconvert them back to their actual usernames when making requests to LDAP Okta (so they're in firstname.lastname@email form).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7437","02/24/2020 14:39:09",2,"Add distro UI option to reprovision Jasper on all hosts on save ""Jasper is automatically restarted when all hosts need new credentials. However, there should be an option to mark all hosts in the distro as needing Jasper to be reprovisioned (e.g. for critical Jasper service updates). The Jasper restart job will have to get the latest curator binary for this. This restart should still have lower priority than the provisioning type conversion jobs. Also, maybe it would be nice to have a restart button on the host page so we don't have to roll over the entire distro host fleet.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7438","02/24/2020 15:24:03",1,"Commit queue misses a nil patch ""If the commit queue tries to process a non-existent patch it will panic instead of dequeueing. We should catch the [nil patch case|https://github.com/evergreen-ci/evergreen/blob/a636d364299d42dd6a710a4f4ece931ea04d50e8/units/commit_queue.go#L273] and probably {code:go}defer dequeue(){code} too.""","defer dequeue()",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7439","02/24/2020 15:26:34",1,"Require rootdir for all Windows provisioning types ""Currently it's not required on legacy provisioned Windows hosts. However, it would greatly reduce the amount of complexity in specifying Windows if we used native paths to files more consistently instead of using Unix-style paths and paths relative to HOME in some (but not all) of the relevant places (excluding binaries, which must be invoked with Unix-style paths).""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7442","02/24/2020 20:27:06",2,"Attempt to give task directory write permissions before removing ""On static hosts, many task directories are not being cleaned up because files/directories in tasks directories are set to read-only permissions. This is bad, because it's currently filling up the disk space on the mac hosts. When the agent starts up, it tries to clean up the whole directory since it has {{--cleanup}}, which can spend a substantial amount of time doing wasteful work (for example, on the latest deploy, a macos-1012 host took more than 15 minutes before it started attempting because it had 32GB of uncleaned task directories that it tried removing). During startup, there is no heartbeat to the app server during this time to tell it that the agent is alive, so after a long period of no communication, the app server might try to deploy again on top of the agent, which is actually alive but stuck in cleanup mode.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7443","02/25/2020 16:41:25",0,"Ensure tasks cannot be put in multiple task queues ""Scheduled tasks should not be able to go into multiple task queues. Otherwise, the scheduler may over-allocate hosts (which it bases off of the primary task queue lengths).""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9135","02/25/2020 16:57:09",2,"Create API user manager from basic user manager ""We need a user manager that can auth in service users to make API requests, but not log into the UI with them. The naive user manager almost does this, but not quite. The source of truth for the special """"service"""" manager should be the pre-populated list of users from settings, not the database. Right now, the naive manager takes some info from the list but also logs in users who exist in the users collection. Essentially, we should have the same manager but only with the ability to use the API (using the pre-populated api key from the settings).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7444","02/25/2020 17:04:04",0,"Fix errors on spawn host page ""There's a lot of errors saying {{$scope.curHostData}} is undefined.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7445","02/25/2020 18:38:48",2,"Update the visibility parameter on the wiki to include the signed option ""Update the wiki. [https://github.com/evergreen-ci/evergreen/wiki/Project-Commands#s3-put |https://github.com/evergreen-ci/evergreen/wiki/Project-Commands#s3-put] Change from: To: Since I would be making this change directly in the UI (as opposed to with a PR), please approve here. I'm therefore changing the status of this ticket to """"in code review"""", even though there is not code. ""","- ``visibility``: one of """"private"""", which allows logged-in users to see the file; """"public"""" (the default), which allows anyone to see the file; or """"none"""", which hides the file from the UI for everybody. visibility: one of """"private"""", which allows logged-in users to see the file; """"public"""" (the default), which allows anyone to see the file; """"none"""", which hides the file from the UI for everybody; or """"signed"""", which displays a pre signed url, allowing logged-in users to see the file (even if it's private on s3). Visibility: signed should not be combined with permissions: public-read or permissions: public-read-write"""". ",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9087","02/25/2020 19:07:47",0,"POST to login instead of GET ""Removing the GET route in cedar will be future work once clients have had sufficient time to update.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9148","02/25/2020 19:09:17",0,"Remove GET route for login ""This should be done after clients have had time to update.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9179","02/25/2020 22:52:14",1,"Run Jasper without resource throttling ""According to [this post|https://apple.stackexchange.com/questions/322557/why-is-an-application-running-so-slowly-when-started-with-launchctl], one possible source of slowness when Jasper processes are run is that they're being throttled as launch daemons is launchd thinks they're background processes. We should try running with {{ProcessType}} set to {{Interactive}} to see if performance improves.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7451","02/26/2020 15:05:55",1,"Include log links for previous executions ""We got a BFG (https://jira.mongodb.org/browse/BFG-480805) for a failure of a task that has 4 executions. The BFG was specifically related to the 3rd execution. When we query the API to get the logs for that execution, they are all null (https://evergreen.mongodb.com/rest/v2/tasks/mongodb_mongo_master_enterprise_ubuntu1804_arm64_sharding_3_enterprise_ubuntu1804_arm64_f83f9dcc22156cdf3c3e16a040914806e2e17cf7_20_02_25_04_29_47?fetch_all_executions=1). It looks like all previous executions have null for the log links. Would it be possible to include the log link in the API?""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7453","02/26/2020 15:44:04",2,"Render placeholder cell filter values when the cell is empty (and the filter uses a default value). ""When using the Perf Baron pages, some of the filter cells have defaults when they are empty (e.g.the create time defaults to 2 weeks ago). When a value is defaulted we should set a placeholder so that it is obvious what filter is being used. ""","",0,1,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-8780","02/26/2020 18:26:21",1,"Add replicaCount, liveness and readiness Probes to selected-tests ""I realized recently we are missing some Kanopy configuration elements in the selected-tests service. As a DAG engineer, I should specify replicaCount to be 2 (it is currently 1), so that the selected-tests service is available during deploys. As a DAG engineer, I should also configure [liveness and readiness probes|https://github.com/10gen/kanopy-docs/blob/29771cc7a3c2cf35e6af2a99f67cab3c8c346ef3/docs/application_configuration.md#configure-k8s-liveness-and-readiness-probes] so that the scheduler is able to understand when new pods are ready to service requests, reducing downtime during deploys. AC: * replicaCount is set to 2 in prod.yml and staging.yml * liveness and readiness probes are configured ""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7455","02/26/2020 18:33:53",0,"Prioritize commit queue patches over normal patches ""Commit queue patches are the most important to schedule quickly and shouldn't block behind ordinary user patch builds.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7458","02/26/2020 20:07:16",2,"Replace DescribeInstances with instance metadata ""We can save on DescribeInstances calls (and avoid API limits) by getting the data about the instance from [instance metadata|https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-instance-metadata.html] from within the agent and posting it back to Evergreen. This would also effectively remove the need for the setup host job to run for user data.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7462","02/26/2020 21:27:06",2,"Add priority to notifications ""We should add a priority field to notifications and sort on it when we pull the next batch of unprocessed notifications. This will help address commit queue notifications not being processed.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7464","02/27/2020 12:00:48",2,"Project Triggers are using the wrong commit ""When a build is triggered from a different project, the commit for the trigger is not used. Instead, an old commit, likely the one that's coupled to the latest commit of the repo for the current project is used. See [this link|https://evergreen.mongodb.com/version/sys_perf_3b279bd75f946b46ad62273a66c3e9d0d6826cc0_81c90ce411305bde898e0a69d2a6479d] and the attached screenshot for as an example. I'm marking this ticket as P2 since it seems to affect the behavior of project triggers across the board. Feel free to reprioritize or ping me if this is a project-specific issue.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7465","02/27/2020 15:42:46",2,"Commit queue should refuse to enqueue an empty patch ""If a user tries to add an empty commit queue item to a queue (such as if they pause the item and then enqueue it later without adding any modules) they should get an error.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7467","02/27/2020 16:36:37",2,"Allow copy_redacted_variables to also copy all variables ""maybe turn it into copy_variables and have an option to """"include_redacted"""" the include redacted option will have the same limitation of needing to be admin of the projects etc. we want to be able to copy all params from one project to another. we take the following steps: 1. get both projects 2. delete variables that are only in the target project 3. call the update project with the json object that is source_project[""""variables""""][""""vars""""] 4. call the copy_redacted_variables api. it would be nicer to just call /project/source/copy_variables/target?copy_redacted=true it would also be nice if in dry run mode it will return a json response that, if ran - would in theory do the copy this way we can dry run - save the result JSON. go over it and see it does what we expect it to do, and then just run update_project with the json file we have. (I realize that since redacted variables are involved, this is maybe not possible)""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7468","02/27/2020 16:57:14",1,"Add baseVersionId to Patch schema type ""There is a bug in the new ui on the patch field: the base commit link in the patch metadata card links to `evergreen.com/version/githash`. It should be `evergreen.com/version/baseVersionId` Change the resolved value of the base commit to be correct. Get the correct base version id by doing ` baseVersion, err := model.VersionFindOne(model.VersionBaseVersionFromPatch(version.Identifier, version.Revision))`""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7473","02/27/2020 21:56:35",1,"Race condition setting commit queue to processing ""From this thread: [https://mongodb.slack.com/archives/C0V896UV8/p1582837378072500] Probably multiple amboy jobs running is what's hitting this race, but we should check that processing is what we expect it to be before setting it and dequeueing/processing items.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7475","02/28/2020 16:16:52",1,"In regression by test trigger, fall back to task failure if all tests pass ""similar to if there are no tests""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7487","03/03/2020 17:53:55",1,"lobster UI should provide access to unmodified json ""There’s no way in the UI to get at the original lines emitted by a test. Need an escape hatch. The raw button gives me the whole log as text with no html markup at all. I can’t see or select line numbers or search it. So essentially all lobster features are unavailable if I want to see the json lines as they were emitted by the unit test.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7488","03/03/2020 18:04:52",2,"Fix flaky auth cypress test ""the issue may be that the local evergreen server is not ready when the auth request is made. example test: [https://dashboard.cypress.io/projects/m45k1x/runs/210/specs]""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7489","03/04/2020 14:24:37",2,"Patch tasks table: fetch new tasks when url params change ""* Fetch with the sortBy and sortDir specified in url * page param is reset to 0 when sorting & filtering changes""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7490","03/04/2020 14:25:42",2,"Patch tasks table: fetch more tasks when scrolling table ""* When I scroll down the table, more tests are fetched * If there are no more tests to fetch then the table stops fetching""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7491","03/04/2020 14:26:25",2,"Patch tasks table: loading & error states ""* ANTD skeleton is displayed when tasks are loading * If there is an error then error message is shown""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8479","03/04/2020 14:36:52",1,"setup_ssh_keys.sh shouldn't overwrite keyfiles ""When setup_ssh_keys.sh is run it writes the contents of GITHUB_PRIVATE_KEY and $GITHUB_PUBLIC_KEY to """"~/.ssh/id_rsa"""" and """"~/.ssh/id_rsa.pub"""". If this script is run in a container then this is fine but if it is run locally, it could potentially cause a user to lose important key files. The script should either use different file names (maybe ~/.ssh/selected-tests, ~/.ssh/selected-tests.pub) or it should not overrwrite existing files. ""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9303","03/04/2020 17:30:02",2,"Evergreen depends on all users to be returned as DBUsers ""We have to add database awareness to the basic user manager because the evergreen middleware won't allow gimlet to return any user except a DBUser. This can be a new wrapper manager with a caching layer that returns only the gimlet.Users that it receives when created.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7493","03/04/2020 20:56:56",2,"Very slow aggregations on the testresults collection ""I noticed the read tickets available were about 20 tickets lower than writes. This appears to very slow aggregations on the testresults collection. A lthough these are not a problem at the moment, if the number of these queries grow, they could potentially slow other queries in the system. We should find out the source of these queries and remove or improve them. I've attached a screenshot of an example query.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7496","03/05/2020 16:15:03",2,"Fix build cache blocked update for display tasks ""[b67f30|https://github.com/evergreen-ci/evergreen/commit/b67f3051319873c983583fed8e3b521884599748] updates blocked dependencies in the build cache. It doesn't take into account that execution tasks aren't part of the build log, and the update fails. MarkEnd returns [here|https://github.com/evergreen-ci/evergreen/blob/55a8bc51f4ceb678a7bf60d73197489529bc51d8/model/task_lifecycle.go#L440] and doesn't complete so the [build cache update for status|https://github.com/evergreen-ci/evergreen/blob/55a8bc51f4ceb678a7bf60d73197489529bc51d8/model/task_lifecycle.go#L451] doesn't run.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7500","03/05/2020 20:36:48",2,"PR commit-queue is setting author to ""evrg-bot-webhook"" ""The selected-tests project has been using commit-queue for a while, but yesterday the author for commits to master started showing up as """"evrg-bot-webhook"""": https://github.com/mongodb/selected-tests/commits/master. The same thing is happening on https://github.com/10gen/build-baron-tools/commits/master. Previously, commits to master made by the PR commit queue set the author of the commit to the author of the PR that was being merged. This made it easier for me to track when my PRs had been merged.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8514","03/06/2020 19:40:41",2,"Loggers seem to hang in tests ""Example: https://evergreen.mongodb.com/task/jasper_windows_test_jasper_patch_13f1b54d555f3b8a04fb9efe7671443fb827bec6_5e629dd47742ae721bc237ea_20_03_06_19_00_37/1""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7511","03/06/2020 21:42:20",1,"Patch page: add commit queue position to metadata ""Query it in GET_PATCH Render it""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7512","03/06/2020 21:43:09",1,"Patch page: add “commitQueuePosition” field to the patch schema type ""How to get commit queue position: [https://github.com/evergreen-ci/evergreen/pull/3453/files#diff-17e8f871e4f3e95fbfafe8e4ccecf48aL76]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7513","03/06/2020 21:43:50",2,"Patch page: build variants card ""* Tasks for each build variant are displayed as cube * TaskCube component * Hovering task shows tooltip with task name * Clicking task goes to task page""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7514","03/06/2020 21:46:11",2,"Patch page: gql query for patch build variants ""Add build variants field to Patch type Build variant type: * Variant name * Tasks * Status * Name * ID""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7516","03/06/2020 21:47:32",1,"Patch page: show task count ""On the /patch route, show the task count (image attached that corresponds to mockups) Designs: [https://mongodb.invisionapp.com/share/S7W848CN3GC#/screens/408295791] A field on the Patch type exists called """"taskCount"""" - it represents the total number of tasks of a patch. Add this field to the Patch query on the /patch route - this value represents the Y in """"X/Y tasks"""" The X in """"X/Y tasks"""" is determined by getting the length value of the tasks array returned from the patchTasks query""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7517","03/06/2020 21:49:24",1,"Patch page: ""search by variant name"" input field ""On change it updates the url using a debounce""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7518","03/06/2020 21:50:17",1,"Patch page: ""search by variant name"" fetch new tasks ""# When url search param changes, fetch new tasks based on search param value # Cancel old requests that have not finished when a new one is made for search""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7519","03/06/2020 21:51:59",1,"Patch page: add variant param to gql resolver ""* Add variant param to patchTasks resolver * In db function, apply match with variant param value""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7520","03/06/2020 21:54:03",1,"Patch page: ""search by task name"" input field ""On change it updates the url using a debounce""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7521","03/06/2020 21:54:39",1,"Patch page: ""search by task name"" fetch new tasks ""* When url search param changes, fetch new tasks based on search param value * Fetch tasks filtered by task input""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7522","03/06/2020 21:55:40",1,"Patch page: add taskName param to gql resolver ""In db function, apply match with variant param value""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7523","03/06/2020 21:57:02",1,"Patch page: filter by status - dropdown menu ""* *Dropdown checkbox menu with all statuses (1)* * Clicking parent of nested statuses selects/unselects all status in that group * Store static list of statuses on the client""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7524","03/06/2020 21:58:28",1,"Patch page: filter by status - clicking on status updates ""statuses"" url param ""When parent of status group is clicked, include all nested statuses in url params""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7525","03/06/2020 21:59:21",2,"Patch page: filter by status - fetch filtered tasks ""When """"statuses"""" url changes, fetch new tasks with new params from url""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7526","03/06/2020 22:00:08",0,"Patch page: filter by base status - dropdown menu ""* Clicking parent of nested statuses selects/unselects all status in that group * Store static list of statuses on the client""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7527","03/06/2020 22:00:47",0,"Patch page: filter by base status - update ""baseStatus"" url param ""When parent of status group is clicked, include all nested statuses in url params""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7528","03/06/2020 22:01:30",2,"Patch page: filter by base status - fetch filtered tasks ""When url param changes, fetch new tasks with new params from url""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7529","03/06/2020 22:37:06",1,"Include cookie in cy.visit header ""Every call to cy.visit causes the auth/cookie to be cleared. Include a cookie in the request so that we do not have to login after every call to cy.visit. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7530","03/06/2020 22:53:37",2,"Task page: dependency resolver for GQL Task type ""task gql type should have a dependency array where each item has the tasks display name, build variant and if it must fail or must finish""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7531","03/06/2020 22:55:33",2,"Task page: spawnHostLink field in Task GQL type ""link to spawn host page, button lives in medadata card""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7532","03/06/2020 22:57:07",2,"Tasks page: Render action buttons in top right of page ""attach on click to mutations. set priority should accept an input. ""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7533","03/06/2020 23:02:18",1,"Task page: Refetch dependsOn field ""should refetch after a certain so the user is up to date without refreshing the page""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7537","03/09/2020 01:13:18",0,"TestsTable should be completely tested in Cypress ""TestsTable is partly tested in cypress and jest. We should rewrite the jest tests in cypress for consistency and also because cypress tests are much easier to understand when something breaks""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7541","03/09/2020 15:39:28",0,"Unattached volumes disappear ""If a volume is created with the [ec2Manager|https://github.com/evergreen-ci/evergreen/blob/c905b1bea0fddfeab8fa10ad0f40a8db8fdc349f/cloud/ec2.go#L1181] and not attached to a host it disappears from the AWS account but is still in Evergreen's db.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8945","03/09/2020 17:29:39",1,"Allow bypassing username/password login and set API keys directly ""In the migration from LDAP to Okta, there will no longer be service user support. Therefore, we have to use the users directly from the database without authenticating them against the Okta backend (so they auth into the API using an API key instead of a password to auth against LDAP).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7545","03/09/2020 18:20:48",1,"Place an Evergreen config file on virtual workstations ""Put the user's credentials in ~/.evergreen.yml on virtual workstation spawnhosts""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7546","03/09/2020 19:23:10",0,"My patches: render header ""Render the """"My Patches"""" header title""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7547","03/09/2020 19:25:11",1,"My patches: search input for patch name ""* Render it * onChange, update “name” url param * When url “name” param changes, fetch new patches""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7548","03/09/2020 19:25:32",1,"My patches: dropdown menu of statuses ""* Render it * onChange, update “status” url param * When url “status” param changes, fetch new patches""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7549","03/09/2020 19:26:09",2,"My patches: commit queue checkbox ""* Render it * Checked by default * onChange, update the url “commitQueue” param * When url “commitQueue” param changes, fetch new patches""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7551","03/09/2020 19:28:11",2,"My patches: PatchCard component ""DOES NOT INCLUDE DROPDOWN (aka the menu elipsis with patch actions) * Polls server for patch status and patch tasks * Stop polling when component unmount * Use useEffect to initialize polling and to unsubscribe * Updates patch status and task statuses in real time""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7552","03/09/2020 19:28:44",2,"My patches: list of patch cards ""Renders a paginated list of PatchCard component""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7554","03/09/2020 19:32:45",2,"My patches: ""Reconfigure tasks/variants"" modal ""When """"Reconfigure tasks/variants"""" is selected from PatchCard dropdown menu, show modal for reconfiguring patch. Designs coming""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7555","03/09/2020 19:38:00",2,"My patches: gql mutation to reconfigure patch ""One of the options from the dropdown menu of the PatchCard is to reconfigure patch. This means changing the variants and tasks that are ran fro the patch. The mutation should take an array of tasks and an array of variants and update the patch with those params. This mutation does not schedule the patch. It only updates the DB model. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7556","03/09/2020 21:52:48",1,"Task page: Display test counts for test table ""display failed test count in test table tab display filteredTestCount/totalTestCount above table""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7557","03/09/2020 21:53:54",1,"Task page: Add count field from taskFiles query ""display the number of files on the files tab""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7558","03/10/2020 01:46:29",2,"Lobster's log substitution of large numbers can lose precision ""BSON has a NumberLong type, a 64-bit integer. The lobster log conversion presumably treats MongoDBs attribute map as JSON which likely parses all numbers into a floating point value. Large 64-bit numbers can then lose precision.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7562","03/10/2020 22:05:36",0,"Add jiraLink field to TaskEventLogData gql type ""this is so we can link to the Jira issue in the event logs in spruce""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7565","03/11/2020 14:38:35",1,"Reduce logging in scheduler ""The scheduler produces a lot of noisy grip logs that make it difficult to find distro logs in Splunk. We should remove the ones that log frequently and don't produce a lot of useful output.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8918","03/11/2020 18:49:13",0,"Sanity check for empty user ""There are many noisy logs for requests being made to evergreen through the user middleware that don't have a user in the headers. The header check in user middleware should check that both the API key and the name are set before trying to get it from the user manager.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7569","03/11/2020 19:17:17",1,"Commit queue sends notifications to the wrong item ""Recently when an item in the queue finishes processing the queue sends a notification to the next item in the queue that hasn't even started processing yet and not for the item that finished processing.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8893","03/11/2020 19:19:50",0,"Check for potentially multiple cookies with the same name ""Staging and prod can be confused because it's not clear which cookie the browser will present first to the request. ""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8651","03/11/2020 19:24:12",1,"Add get_version to Build ""As a DAG engineer, I'd like to add get_version method to Build objects, such that, I can get a build's version using evergreen.py. AC: * get_version added to Build class.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8542","03/11/2020 20:31:10",2,"Allow aborting amboy jobs by prefix ""Currently you can abort jobs by id, or you can abort all jobs, but you can't abort only one kind of jobs. We should add this functionality, possibly by allowing aborting jobs by prefix.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7572","03/11/2020 22:04:05",2,"Logs url from API is wrong ""The Build Baron worker occasionally fails to download logs to search for failures. Looking into one of these failures, it seems like we are getting the wrong log URL from the API. It does seem to occur when there have been multiple executions of the task and we are looking at an execution that may not be the latest. For example, [This task|https://evergreen.mongodb.com/rest/v2/tasks/mongodb_mongo_v4.4_enterprise_debian10_64_jstestfuzz_concurrent_replication_continuous_stepdown_flow_control_09_enterprise_debian10_64_340ddb282a92bb5e97d0c43b7d44c26ec7d8da0e_20_03_10_13_18_27?fetch_all_executions=1]. Looking at the first executions, all the logs seem to append an extra """"_0"""" to the task id in the URL, which makes the links invalid. The log to the system log is: https://evergreen.mongodb.com/task_log_raw/mongodb_mongo_v4.4_enterprise_debian10_64_jstestfuzz_concurrent_replication_continuous_stepdown_flow_control_09_enterprise_debian10_64_340ddb282a92bb5e97d0c43b7d44c26ec7d8da0e_20_03_10_13_18_27_0/0?type=S""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7577","03/12/2020 01:45:41",1,"Log event when host allocator changes static host status ""This is convenient for tracking events such as transitioning between running and quarantined/terminated.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8370","03/12/2020 13:53:37",2,"Double slashes and permanently moved errors ""See [here|https://mongodb.splunkcloud.com/en-US/app/search/search?earliest=-4h%40m&latest=now&q=search%20index%3Dcedar%20auth&display.page.search.mode=smart&dispatch.sample_ratio=1&display.page.search.tab=events&display.general.type=events&display.visualizations.charting.chart=line&sid=1584021158.2338196] for more information. This may be due to redirects.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9125","03/12/2020 14:19:31",0,"Increase repobuilder timeout ""The new repobuilding service has to build more repos, so it will hit the 1 hour timeout and fail to sync with pail.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8683","03/12/2020 14:45:31",2,"Update selected-tests to use poetry ""Update selected-tests to use poetry""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7580","03/13/2020 15:29:22",1,"Add AWS credentials for S3 push to agent ""The agent needs the credentials to push task directories to S3. This should probably go in the environment for the agent setup.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7581","03/13/2020 16:43:29",0,"Add newline for distro aliases listed in CLI ""It needs whitespace or the first alias is printed on the same line.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7583","03/13/2020 18:37:12",2,"Lobster json pretty printing goes away with scroll ""Scroll around the one bookmark here with the json pretty printing on to understand: https://logkeeper.mongodb.org/lobster/build/56e19e8ab9e7042a85b559a81a5be69b/test/5e6ab05c9041301fc71f79bd#bookmarks=53211&l=1""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7641","03/16/2020 02:26:41",2,"So many log lines of ""Attach test logs succeeded"" in Java tests ""It's coming from For example, see https://evergreen.mongodb.com/task/mongo_java_driver_tests_zlib_compression__version~3.6_os~linux_topology~standalone_auth~noauth_ssl~nossl_compressor~zlib_jdk~jdk8_test_cedb69c82bd4bd321a2929b4110f574dee5a2881_20_03_13_16_01_44. I think this is because in Java each Junit test generates its own file, and the command is logging every single one. Anything we can do to dial down the verbosity of this logging?"""," """"upload test results"""": - command: attach.xunit_results params: file: ./src/*/build/test-results/*/TEST-*.xml ",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7585","03/16/2020 16:47:58",1,"Do not show validation warnings for distro aliases ""If you run {{evergreen validate}} on a config containing a distro alias for distros/run_on, it will give warnings about a nonexistent distro even though the distro alias exists. It shouldn't give warnings for using a distro alias.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7586","03/16/2020 17:52:31",1,"Download working directory from spawn from task ""The linked tickets describe preserving and restoring the working directory from tasks. Another use case is restoring the directory to a spawn host in order to debug a failure.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7588","03/16/2020 20:53:27",1,"Shut down agent before AWS shuts down spot instance ""Poll this endpoint on AWS hosts https://aws.amazon.com/blogs/aws/new-ec2-spot-instance-termination-notices/ and exit the agent if the endpoint returns that the host is going to be terminated. This should ensure that the agent does not report a failed or system failed status to the app server. Instead we will rely on the app server cleaning up this task to mark it system unresponsive.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7589","03/16/2020 20:54:17",1,"Do not create tickets for system-unresponsive tasks ""System-unresponsive tasks do not require human investigation, so we should not create tickets for them. Only setup, system, and failed tasks should create tickets.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7590","03/16/2020 20:57:08",0,"Noop new agent downloads from monitor if instance is going to shut down ""On an AWS host, if the instance is going to shut down, the monitor should not download a new agent. The agent has likely exited due to detecting that the instance is going to shut down. Instead of exiting the monitor, simply noop the download and setup, since having code that exits the monitor is potentially risky.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7593","03/17/2020 13:31:42",1,"Handle distro aliases in host.create ""Currently there's no handling of distro aliases in the host.create command.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7595","03/17/2020 17:35:14",1,"Decommission idle hosts of deleted distros ""The idle hosts query relies on finding distros that exist. However, for hosts whose distros have been deleted (e.g. for distro aliases), these hosts also have to be decommissioned.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7596","03/17/2020 19:25:14",2,"A display task was finished before it actually finished ""A display task had two execution tasks that timed out. We logged an event for the display task finishing before the second execution task timed out.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7597","03/17/2020 19:33:32",2,"Display task BFGs should link to failing execution tasks' logs ""Since [EVG-7304|https://jira.mongodb.org/browse/EVG-7304] a single BFG is created for all test failures in a display task. We should change the task log links to point to execution tasks' logs since the display task has no logs.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7599","03/18/2020 14:37:49",0,"Handle distro aliases when checking distro permission scopes ""When distros are deleted, the distro IDs used to check permissions should go to their aliased distro instead of the deleted distro.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7600","03/18/2020 15:53:53",2,"Investigate finding distros by ID in all usages ""A bunch of calls to find distros by ID should probably support usage of aliases, but they have to be handled on a case-by-case basis.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7601","03/18/2020 16:20:29",2,"Handle AZs that don't support the instance type in Fleet requests ""Some AZs don't support all instance types. If a distro's instance type isn't supported by an AZ we should exclude that AZ from Fleet requests (by way of its subnet ID) for instances of that distro, or at least catch the error and not retry. One possibility is to get the list of supported instance types for an AZ [through the API|https://docs.aws.amazon.com/sdk-for-go/api/service/ec2/#EC2.DescribeInstanceTypeOfferings] (and probably cache it somewhere)""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7602","03/18/2020 18:09:46",1,"Intent host IDs sent to DescribeInstances API ""This is a problem, because it causes Evergreen to exceed its API limit, which prevents it performing other operations. Evergreen did this 5000 times in a 1-hour period https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Devergreen%20client%3D%22%5C*cloud.awsClientImpl%22%20message%3D%22AWS%20API%20call%22%20metadata.level%3D70%20%20api_name%3DDescribeInstances%20InvalidInstanceID.Malformed&earliest=1584547200&latest=1584550800&display.page.search.mode=verbose&dispatch.sample_ratio=1&sid=1584554969.4654267.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7603","03/18/2020 20:55:14",1,"Virtual workstation home volume device name isn't set ""When we attach a volume to the instance we should pass the distro's home volume device name to AWS.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7604","03/19/2020 00:54:55",2,"Patch page: schedule patch tasks mutation ""The schedule button on the patch page calls modify version to schedule all tasks in legacy UI, clicking """"schedule all tasks"""" on patch page makes a PUT request to [/version/\{versionId}|https://evergreen.mongodb.com/version/5e727e7932f41772e1594452] with a req payload of """," {active: true, abort: false, action: """"set_active""""}active: trueabort: falseaction: """"set_active""""} ",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7606","03/19/2020 14:34:09",2,"Spawn host display names ""Add a display name to spawn hosts that user's can change to something more memorable than instance ids.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7607","03/19/2020 14:47:41",2,"sys-perf benchmarks return ""could not find task on base commit"" ""sys-perf has started using generate.tasks, which appears to interact poorly with historical data. This link https://evergreen.mongodb.com/plugin/json/history/sys_perf_linux_standalone_validate_cmd_patch_aef7cbf49c396df95c02606a9e94d26c892e53ed_5e72767c1e2d17595813fb53_20_03_18_19_52_18/perf returned """"could not find task on base commit"""", presumably because the task didn't exist at all in the base commit. We should see if it's possible to have the logic to look back past base commits even if the task doesn't exist.""","",0,0,1,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7608","03/19/2020 16:06:55",1,"Loading skeleton for task page tabs ""add the loading skeleton for the task page tabs""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7609","03/19/2020 19:05:11",2,"Problems in performance graph displays ""Some graphs are showing artifacts and may be trying to connect the results from different thread levels. ""","",0,0,1,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-9221","03/19/2020 19:06:35",2,"use go1.13 for cedar ""can we deploy from artifacts compiled by computers that aren't julian's laptop. please verify""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7611","03/19/2020 20:24:43",0,"Task table shouldn't push default query params to URL ""currently the task table component pushes default query params to the URL if invalid query param values for the task table. The task table should not push default query params to the URL and infer the correct parameters internally""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7612","03/19/2020 21:19:11",0,"Omit Jasper splunk credentials for virtual workstations ""Virtual workstation hosts (and probably spawn hosts in general) shouldn't get these credentials.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7614","03/19/2020 21:35:51",0,"Validate that user does not spawn a virtual workstation with task data ""Since virtual workstations are just slightly different spawn hosts and do not have a github key to clone repositories, it would be invalid to load task data. It is still possible to do this via the UI or just a straight rest request.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7616","03/20/2020 13:40:12",0,"Add environment variable option to skip integration tests ""Almost 90% of the time I don't care about the integration tests and I don't want to populate the credentials, so I just want to skip them.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7617","03/20/2020 14:50:12",1,"Make url query params between patch and task pages the same ""The query params used between patch and task page should be the same. the params passed to the queries should also be the same (where there is overlap). This specifically relates to the sortDir param, which has a value of ASC || DESC on the patch page and 1 || -1 on the task page. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7620","03/20/2020 16:01:14",1,"App crashes when switching back and forth between tabs ""Attached are screenshots of the error on the patch and task pages. the task page is the screenshot with the 422 error. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-9065","03/20/2020 21:13:08",1,"cleanup paths for temporary directories ""currently we hardcode some stuff in data, but it should be app-level configurable""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7622","03/20/2020 21:16:38",1,"cron buildvariant has not run ""I used cron syntax for a buildvariant to run OCSP tests. The tests are supposed to run every 14 days and the buildvariant first shows up in the waterfall on 2/28/2020 but all of the runs show that it is inactive. Is the syntax incorrect? {{evergreen validate}} does not report any errors. Link to the section of the config file that defines this buildvariant: [https://github.com/mongodb/mongo-go-driver/blob/master/.evergreen/config.yml#L1605]""","",0,0,0,0,0,1,0,0,0,1,0,0,0,0,0
+"EVG-7624","03/21/2020 14:26:16",2,"Add regex patterns to project to determine what tasks should run ""Add to UI, as well as CLI. Should handle variant regexes, task regexes, variant tags, and task tags.""","",1,0,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7625","03/21/2020 14:26:41",2,"Add authorized github usernames to project ""Only these users will be able to create versions when they push git tags""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7626","03/21/2020 14:27:02",2,"Add git_tags_only tag to tasks in yaml ""Introduce a """"git_tags_only"""": true setting for tasks in the project yaml that should only run when a git tag is pushed (i.e. not in mainline or patch builds) and the commit author is part of the allowed github usernames""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7628","03/21/2020 14:27:34",1,"Document process ""Document how to set this up to simplify setup for users""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7629","03/23/2020 12:49:48",0,"Add Makespan field to GQL Schema ""Add Makespan field to GQL schema to show in the Task metadata panel""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7630","03/23/2020 13:18:27",1,"Alias static hosts with build/test suffix ""We have to migrate all the static hosts to use the large/small convention instead. The old host task queues should be cleared after the distro is aliased and deleted to avoid duplicate task notifications, since their queues tend to be very long.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7631","03/23/2020 14:13:39",2,"Patch page: map patch task status & base status to corresponding badge color ""The status and base status columns in the patch tasks table should show badges that correspond to the status color Will require putting new tasks in DB for patch that cover all available patch statuses [https://mongodb.invisionapp.com/d/main/?redirHash=#/console/19479418/408295791/preview?scrollOffset=229]""","",1,0,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7632","03/23/2020 14:31:09",1,"Git commit date can be later for modules than for the main repo ""The commit queue pushes module commits before the main repo so the module commit hashes recorded in the main commit's manifest will be accurate. Despite this, sometimes the main repo commit's commit date precedes the modules' and the manifest gets the wrong commit.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7633","03/23/2020 14:49:40",1,"Task Page: Add file count to GQL schema ""this is necessary to show the number of files in the files tab.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7634","03/23/2020 14:51:28",1,"Task Page: Add Test Count to GQL schema ""task type should return the failed test count to display in the tests tab taskTests mutation should return total number of tests in task tests taskTests mutation should return number of tests found from filters excluding page and limit ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7635","03/23/2020 16:12:12",1,"Do not terminate user data hosts that pass provisioning cutoff if they are running a task ""Due to a job backup today, some hosts were stuck in starting but since they ran with user data, the agent managed to pick up tasks even though they were not marked as provisioning. We should let these hosts continue as long as they are running tasks.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9181","03/23/2020 20:34:24",2,"split options for deleteOnSync ""add separate options for delete on sync pull-only/push-push (and make the current option just set the other two option, with validator logic to prevent ambiguous settings.) ""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7644","03/24/2020 18:46:43",0,"Allow status to be set on terminated static host and preserve quarantined status ""Static hosts are terminated if removed from a static distro, but should be set to running if moved to a different distro.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7645","03/24/2020 19:44:31",2,"Report relevant errors to bugsnag ""report errors to bugsnag and new relic [https://docs.google.com/document/d/1G7Fyp6VwCIwfY2kwzEUKRnQuYD1yAV3wb5Tuv3GF_Xg/edit] * this should be high alert to bugsnag: [https://github.com/evergreen-ci/spruce/pull/106/files/9fcc28747dd1da382d4c499defe56c83d47eb1a1#diff-a7ecbc3c8ce6748c1957900fcbc52ae3R123]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8408","03/24/2020 21:19:46",1,"Poplar gRPC doesn't set timestamps ""Events added through poplar gRPC don't seem to have their timestamps reflected in the resultant ftdc files. It looks like the [Export|https://github.com/evergreen-ci/poplar/blob/3ed14b6c043a6f2d434241a302690e904daa47ff/rpc/internal/collector.ext.go#L8] function doesn't pass the timestamp field on to the [Performance|https://github.com/mongodb/ftdc/blob/0659acdc6e9ea8863e0d9e888fb09da4287343d2/events/performance.go#L26] struct.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7647","03/25/2020 13:42:31",2,"Lobster UI lag when selecting options or highlighting log line ""The latest [version|https://github.com/evergreen-ci/lobster/commit/9968954d3568ff4899f0e807e6eb0a6054d5a5f3] of Lobster appears to have a noticeable lag between the time of executing an action like clicking an option button or highlighting a log line and the action having an effect. This [screencast|https://www.dropbox.com/s/w5hr9e41ckltfn9/lobster_ui_slow.mov?dl=0] tries to capture the effect. This happens on Chrome (80.0.3987.132) and Safari.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8955","03/25/2020 14:05:12",1,"Standardize cronjob names in selected tests ""As a DAG engineer, I want all the selected tests cronjobs to have the same prefix, So that I can easily detect them in the kanopy output. ---- AC: * all cronjobs in selected tests start with 'selected-tests'""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7649","03/25/2020 14:51:04",1,"fix patch-route flakey test: Fetches sorted tasks when table sort headers are clicked ""right now this test is muted""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-9202","03/25/2020 17:01:05",1,"Better REST logging ""Many of the admin commands have bad service side logging so it makes it hard to figure out issues when something goes wrong. We should have more robust logging in these cases.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7651","03/25/2020 17:55:00",2,"Some json objects in lobster aren't expanded ""In https://logkeeper.mongodb.org/lobster/build/b0b5735508d54fe614397bd9d310f5be/test/5e7b8263f84ae831d1166f42#bookmarks=0%2C478%2C3498%2C21167%2C21405%2C21432%2C21442%2C21468%2C21469%2C21471%2C21472%2C21473%2C21474%2C21619%2C21707%2C21728%2C21828&f~=000~%2Fmongod&f~=100~config.shards&l=1, Line 3498 is not expanded when the """"pretty print"""" setting is enabled.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7653","03/25/2020 18:39:40",0,"Incorrect trend graph representation in some cases ""Incorrect trend graph representation in some cases [aggregation on Standalone Linux inMemory|https://evergreen.mongodb.com/task/performance_4.4_linux_wt_standalone_aggregation_02e12f2b4acfb8f1d401ed78a04fe4b6e23b9976_20_03_20_20_37_41##Aggregation.Group.TenGroupsWithSumJs] When selected commit is on top of improvement then the graph is built incorrectly: !Screen Shot 2020-03-25 at 11.31.36 AM.png|width=600! When the selected commit set up on the any other point (other then “first failing revision” point) then graphs look good: !Screen Shot 2020-03-25 at 11.32.23 AM.png|width=600! ""","",0,0,1,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7654","03/25/2020 18:45:24",2,"Remove binary representation of generate tasks ""We currently save both a binary version and a string version of generate tasks. We have been running with the string version for some time, so we're confident that it works. We should stop persisting the binary version.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7656","03/26/2020 00:38:28",2,"enable extended amboy stats in evergreen ""there's a constant in the code to disable it, and the goal is to flip the query, but we just have to make sure that the operations aren't going to be too expensive (we've tried it in the past), but I think we've added indexes so it should be fine, but it's a bug that doesn't reproduce in staging""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7657","03/26/2020 11:24:46",0,"Do not pass host's SSH options into SSH command runners ""Since the SSH options are already available without needing to pass them in, it doesn't make sense to pass a property of the host into a host method.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8426","03/27/2020 01:04:01",1,"ssh keys not being created in cronjobs ""The ssh keys for test selection workers are not getting created properly and causing the scripts to fail. It looks like this is due to attempting to pass an argument when we source the script to create the ssh keys, which doesn't work. See [here|https://github.com/mongodb/selected-tests/blob/fe60908fcd767d40b285dd854659ca9f7a16f2fb/cronjobs/update_task_mappings.sh#L5]""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7662","03/27/2020 12:25:57",1,"Select first matching distro from spawn host page if distro is aliased ""If you try to spawn a host from the task page for a distro that has been deleted, just try to find the first matching distro which has an alias (or give some kind of error message saying no such distro exists).""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7666","03/27/2020 14:37:10",1,"Do not call DescribeInstances twice to get volume total size ""When we call {{cacheHostData}}, we already have the instance info from {{DescribeInstances}}. However, calling {{getVolumeSize}} within that function makes another request to {{DescribeInstances}}, so we unnecessarily double the amount of {{DescribeInstance}} calls we need to make for caching. We do this every time we cache the host data (e.g. during the cloud host ready job).""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8424","03/27/2020 14:42:26",2,"Why does an ID of 0 lead to a monotonically-increasing value? ""Would be a question ticket, but it looks like that type doesn't exist for this project. Was just wondering what the reasoning is behind [this|https://github.com/mongodb/ftdc/blob/master/events/performance.go#L169] if-statement that leads to IDs monotonically increasing if they're at zero. It's preventing us from zero-indexing threads and I was curious what the rationale was / whether there are cases when we should just let it monotonically increase.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7668","03/27/2020 15:31:14",2,"Investigate Fleet error message ""When we tried starting an arm instance with Fleet we got an error message which indicates that instead of asking for an instance in us-east-1a we should ask for an instance in us-east-1b or us-east-1d. The request overrides (included in the log message) included all three, though us-east-1a was first. The default subnet for the distro (included in the launch template) could have been us-east-1a as well. In general, investigate whether AWS is respecting all the overrides, or just picking the first.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7670","03/27/2020 17:42:01",1,"Spawn host queries for host provisioning should check UserHost ""Some places check for UserHost, but other places check StartedBy, which is unreliable because it could include host.create hosts.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7671","03/27/2020 18:52:36",1,"Intent hosts for hosts started by tasks are removed after three minutes ""The query the host allocator uses to find stale host intents also includes stale intents for hosts started by tasks and removing those is the purview of [PopulateHostTerminationJobs|https://github.com/evergreen-ci/evergreen/blob/49d6d438cb3c5aed7f64f01deef2832aafd374c0/units/crons.go#L271] ""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7674","03/30/2020 03:55:59",1,"Allow -y argument to evergreen commit-queue with multiple commits ""With multiple commits on the current branch, run the following: {{evergreen commit-queue merge -p mongodb-mongo-v4.4 -y}} Expected: The commits are submitted to the merge queue. Actual: Evergreen prompts me: {{Commit queue patch has multiple commits. Continue? (y/n):}} I think that """"-y"""" should override all yes/no prompts.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7679","03/30/2020 17:28:58",1,"Write docs for s3.push and s3.pull ""These new commands should be documented in the wiki. There should be lots of strong words about the potential exposure of secrets in the task directory. Also document the {{variants}} field for commands, dependencies, and requirements. Also document how to use the CLI options.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7680","03/30/2020 18:03:50",2,"Add project-level configuration to enable s3 push/pull via project config and patching ""Since these commands can expose secrets stored in the agent's task directory and anyone in evergreen can run a patch whose task directory can be later pulled, project admins should opt into this feature. Alternatively, there could be an option to allow s3.push and s3.pull but not allow {{evergreen fetch}} to download the task dump. This would keep the task directory within the agent's host.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7681","03/30/2020 19:25:40",0,"Task groups not executing sequentially ""We're seeing build failures again where the """"stitch_support_run_tests"""" task starts executing before the """"stitch_support_install_tests"""" finishes. As a result the former task fails, because the test binary it wants to run isn't finished building and installing.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7682","03/30/2020 19:33:24",1,"Use order field instead of patchNumber from Task gql type & rm patchNumber field ""FE: patch number should accessed from the order field on the Task type instead of patchNumber on the Task type. BE: Remove patchNumber from the Task type""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7684","03/30/2020 22:56:38",2,"Add UI option to pull data from this task sync from s3 into a spawn host ""If this task's working directory was synced with s3 (through a CLI request to do so), it should have a UI display to show the command to pull from the task sync directory. ""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7685","03/30/2020 23:08:55",1,"Fix Create Volume device name generating ""Caused the 3/30 outage""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7686","03/31/2020 14:36:01",2,"Performance Baron sorting by revision should sort chronologically ""As a Build Baron, When I sort by revision, The change point data should appear in chronological order. ---- AC: * the revision field is essentially random so sorting by revision whithin a page is meaningless and not helpful * there is a one to one mapping from the revision to the order field. The revision sort should use the order field to allow sensible sorting, as the Order field is the monotonically increasing field that essentially represents the create time of the performance data.""","",0,1,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7687","03/31/2020 14:44:44",3,"Performance Baron number type filter should support abs ""As a Build Baron, I should be able to filter by absolute value (e.g. '>= abs(0.05)'), So that I can filter change points that would generally be viewed as noise. ---- AC: * In addition to the currently supported operators ('>','>=','<','<=' and '==') number field types should also support abs() to filter by absolute value. * the default filter for the hazard field should be set to '>= abs(0.05)'""","",0,1,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7688","03/31/2020 15:54:38",2,"Limit GraphQL playground mutations to superusers ""Queries are already limited to authenticated users. The goal of this ticket is to limit mutations.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7689","03/31/2020 16:17:39",1,"Ignore extra variables in project validation ""A lot of projects give this error: {{WARNING: yaml: unmarshal errors: line 1: field variables not found in struct model.ParserProject}} This is because many projects use variables. There's also a project validation error message missing a value for the format string on [this line|https://github.com/evergreen-ci/evergreen/blob/f64dfedbd176edf4959711fe26b0bfe0a41ffe43/operations/validate.go#L98].""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7691","03/31/2020 17:47:43",0,"TreeSelect width should be dynamic / 100% of its parent ""I made a temporary fix so that TreeSelect accepts a width prop. However it causes the dropdown div to be less width than its parent. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7693","03/31/2020 22:44:18",1,"Don't use all caps for status badge labels ""They should be in title case like the rest of the labels""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7694","03/31/2020 23:24:22",2,"Change font and add plugins ""change font to consolas/inconsolata plugins: - JavaScript (for server tests) ([EsLint|https://marketplace.visualstudio.com/items?itemName=dbaeumer.vscode-eslint], [Babel|https://marketplace.visualstudio.com/items?itemName=mgmcdermott.vscode-language-babel], [Snippets|https://marketplace.visualstudio.com/items?itemName=xabikos.JavaScriptSnippets], [Booster|https://marketplace.visualstudio.com/items?itemName=sburg.vscode-javascript-booster]) - Golang (for us as well as enterprise tools) ([Go|https://marketplace.visualstudio.com/items?itemName=ms-vscode.Go]) - Snippet/Expansions ([vscode-snippet|https://marketplace.visualstudio.com/items?itemName=vscode-snippet.Snippet]) - Yaml ([YAML|https://marketplace.visualstudio.com/items?itemName=redhat.vscode-yaml]) - markdown/restructured text ([Markdown All in One|https://marketplace.visualstudio.com/items?itemName=yzhang.markdown-all-in-one]) - Bash/shell scripting ([Bash Debug|https://marketplace.visualstudio.com/items?itemName=rogalmic.bash-debug] and [Bash IDE|https://marketplace.visualstudio.com/items?itemName=mads-hartmann.bash-ide-vscode]) - The vim emulation mode ([VSCodeVim |https://marketplace.visualstudio.com/items?itemName=vscodevim.vim]) - Emacs emulation mode ([Emacs Friendly Keymap|https://marketplace.visualstudio.com/items?itemName=lfs.vscode-emacs-friendly]) - Some good themes ([Base 16 Themes|https://marketplace.visualstudio.com/items?itemName=AndrsDC.base16-themes]) - REST ([REST Client|https://marketplace.visualstudio.com/items?itemName=humao.rest-client])""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8462","04/01/2020 15:08:46",0,"DispatchBy should query for tasks before current time ""If it's before the DispatchBy time, we can still dispatch the job. If it's past the DispatchBy time, we should not dispatch the job.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8704","04/01/2020 16:23:17",2,"Add percent_change to change points ""As a DAG engineer, I'd like cedar to calculate percent_change for change points being detected, such that, it can be used for filtering/sorting. AC: * percent_change is calculated and persisted during change point calculation jobs * Migrate existing change points in cedar to have these""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8595","04/01/2020 16:27:07",2,"Add triage section to change points in cedar ""As a DAG engineer, I'd like to add a triage section to change points being detected in cedar, such that, this triage section can be integrated with further triaging functionality. AC: * Triage section added and persisted in default untriaged status for all change points being created * Add default to all old change points See technical design of associated epic for format.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9016","04/01/2020 16:46:13",3,"Restrict calculation of change points to the portion of the time series after the latest triaged change point ""AS a DAG engineer, I'd like to restrict the calculation of change points to the portion of the time series after the latest triaged change point, such that, we do not evict triaged change points and do not constantly recalculate change points over old performance results. AC: * change points only detected for the part of the performance result history that has not yet been triaged""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8557","04/01/2020 16:51:15",2,"Add triage endpoint to cedar ""As a DAG engineer, I'd like to add a triaging endpoint to cedar, such that, change points can be triaged. AC: * Triage endpoint added. Please see technical design for exact API to implement.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9226","04/01/2020 16:59:24",2,"Scoped jobs should be resilient to server restarts ""Repobuilder jobs that are in progress get stuck in progress when the app server is restarted. If they are stuck in progress, the stuck job will block any future job with the same scope from running. They should probably restart if they're stuck in progress. This is likely due to some problematic interaction between scopes, the lock timeout, and job dispatching.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8748","04/01/2020 17:06:43",3,"Add change point list grouped by version endpoint to cedar ""As a DAG engineer, I'd like to add route to get change points grouped by version, such that, we can get a list of change points grouped by version for other components to use. AC: * paginated * conforms to technical design API (page, page_size, total_pages -- do not worry about filtering and sorting for this ticket)""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8602","04/01/2020 17:09:34",3,"Add filtering by variant, task, test, version, measurement, thread_level to get change points by version API ""As a DAG engineer, I'd like to be able to filter change points by variant, task, test, version, measurement, thread_level, such that, these lists can be filtered. AC: * these lists can be filtered using a regex for the above properties Please see technical design for API to implement.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8561","04/01/2020 17:15:23",2,"Add sorting for percent_change and calculated_on to cedar change points lists ""As a DAG engineer, I'd like to be able to sort change points lists from cedar by percent_change, and calculated_on, such that, these lists can be sorted by these properties. AC: * Can sort on these properties""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7699","04/01/2020 17:38:07",3,"Add change point grouped by version table to legacy UI ""As a DAG engineer, I'd like to add a table component to display change points grouped by version to the legacy UI, and expose it as a page in evergreen, such that, people can see expanded metrics change points in evergreen. AC: * Table page available in evergreen * version should be able to be expanded showing change points it contains * Table should allow you to multiselect change points/versions * Existing perf baron page has option to switch to expanded metric baron page * This now page has option to switch to legacy perf baron page * API providing data should be exposed as a service, not actuated directly by table component Note: no filtering/sorting/actions for this ticket Please consult design doc and legacy perf baron page for guidance.""","",1,0,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7700","04/01/2020 17:41:45",3,"Component for filtering by variant, task, test, version, measurement, thread_level when getting change point lists from legacy UI ""As a DAG engineer, I'd like to have a component that allows user to specify filter for change points by variant, task, test, version, measurement, thread_level, such that, these lists can be filtered on the front end. AC: * Component should handle creating basic regexes * Component should allow user to specify an exact regex, either directly, or using some kind of advanced mode * -Component should provide a tooltip describing what it does somewhere, in case user is confused- * Component should be integrated into change point table for each of the above properties * Service updated to pass these arguments through to API calls appropriately * measurement filter should default to the following three measurements: [“95th percentile latency”, “50th percentile latency”, “average read latency”] Here is the [wireframe|https://docs.google.com/spreadsheets/d/1Dh-C2CB15lb7veXP1uoXZNh43MMvZuvx4hX-4mVilSY] for the UI. Here is the [technical design|https://docs.google.com/document/d/1wKLZ-IWjA8-kWAO44tHotzuYei0y2rNX7OQjrKGEo34] for this project. The specific route this component should be backed by is: """," GET /perf/project//change_points_by_version ",1,0,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7701","04/01/2020 17:49:39",2,"Add sorting capability to component for filtering variant, task, test, version, measurement, thread_level ""As a DAG engineer, I'd like to add the ability to specify a sort in the component that handles filtering for the above properties in the change point for expanded metrics list in spruce, such that, we can also sort on these properties. AC: * Can specify a sort order as well * Service updated to pass sort through to API Here is the [wireframe|https://docs.google.com/spreadsheets/d/1Dh-C2CB15lb7veXP1uoXZNh43MMvZuvx4hX-4mVilSY] for the UI. Here is the [technical design|https://docs.google.com/document/d/1wKLZ-IWjA8-kWAO44tHotzuYei0y2rNX7OQjrKGEo34] for this project. The specific route this component should be backed by is: """," GET /perf/project//change_points_by_version ",1,0,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7702","04/01/2020 17:51:46",2,"Component for filtering by percent_change and calculated_on in the Expanded Metrics Perf Baron UI ""As a DAG engineer, In the [Expanded Metrics Perf Baron UI|https://evergreen.mongodb.com/perf-bb/expanded-metrics-change-points], I'd like to have a component that allows user to specify filter for change points by percent_change, and calculated_on, such that, these lists can be filtered on the front end. Percent change and calculated on are floats and datetimes, respectively, and this component should allow users to specify windows for these filters per the [wireframe|https://docs.google.com/spreadsheets/d/1Dh-C2CB15lb7veXP1uoXZNh43MMvZuvx4hX-4mVilSY]. AC: * Component should be integrated into change point table for each of the above properties * Service updated to pass these arguments through to API calls appropriately Note: [Technical Design|https://docs.google.com/document/d/1wKLZ-IWjA8-kWAO44tHotzuYei0y2rNX7OQjrKGEo34/edit#]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7703","04/01/2020 17:55:04",3,"Add sorting capability to component for filtering percent_change, and calculated_on ""As a DAG engineer, I'd like to add the ability to specify a sort in the component that handles filtering for the above properties in the change point for expanded metrics list in spruce, such that, we can also sort on these properties. AC: * Can specify a sort order as well * Service updated to pass sort through to API Here is the [wireframe|https://docs.google.com/spreadsheets/d/1Dh-C2CB15lb7veXP1uoXZNh43MMvZuvx4hX-4mVilSY] for the UI. Here is the [technical design|https://docs.google.com/document/d/1wKLZ-IWjA8-kWAO44tHotzuYei0y2rNX7OQjrKGEo34] for this project. The specific route this component should be backed by is: """," GET /perf/project//change_points_by_version?triage_status_regex= ",1,0,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-9112","04/01/2020 18:02:08",0,"Add option to verbose log the files being downloaded from pull ""This is mostly a UX benefit for users running the CLI pull command, which could take a long time depending on your network bandwidth. In lieu of having a progress bar, it would be nice to give an option to print what file is currently being downloaded from S3.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7706","04/01/2020 18:35:56",2,"Add component for filtering expanded metrics change points by triage status ""As a DAG engineer, I'd like to add a UI component for filtering expanded metrics change points based on triage status, such that, build barons can focus on untriaged change points. AC: * Add and integrate component allowing users to select triage status filter on expanded metrics triaging page * Update service layer to pass through triage filter * defaults to """"untriaged"""" Here is the [wireframe|https://docs.google.com/spreadsheets/d/1Dh-C2CB15lb7veXP1uoXZNh43MMvZuvx4hX-4mVilSY] for the UI. Here is the [technical design|https://docs.google.com/document/d/1wKLZ-IWjA8-kWAO44tHotzuYei0y2rNX7OQjrKGEo34] for this project. The specific route this component should be backed by is: """," GET /perf/project//change_points_by_version?triage_status_regex= ",1,0,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7708","04/01/2020 22:14:18",2,"Allow users to specify region in s3 commands ""As we have more remote users, we should allow buckets outside of the """"us-east-1"""" region to be usable with our s3 commands. Doing this would require letting users specify the region of their bucket in the s3 command.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7710","04/02/2020 14:42:45",1,"Panic in s3.get on windows "" """," {""""message"""":""""hit panic; adding error"""",""""metadata"""":{""""level"""":90,""""time"""":""""0001-01-01T00:00:00Z""""},""""operation"""":""""problem running command 's3.get'"""",""""panic"""":""""runtime error: invalid memory address or nil pointer dereference"""",""""stack"""":[{""""function"""":""""github.com/evergreen-ci/evergreen/agent.(*Agent).runCommandSet.func2.1"""",""""file"""":""""/data/mci/8a4e334fe1089b7dbac8d2778c8083f8/gopath/src/github.com/evergreen-ci/evergreen/agent/command.go"""",""""line"""":115},{""""function"""":""""runtime.gopanic"""",""""file"""":""""/opt/golang/go1.13/src/runtime/panic.go"""",""""line"""":679},{""""function"""":""""runtime.panicmem"""",""""file"""":""""/opt/golang/go1.13/src/runtime/panic.go"""",""""line"""":199},{""""function"""":""""runtime.sigpanic"""",""""file"""":""""/opt/golang/go1.13/src/runtime/signal_windows.go"""",""""line"""":233},{""""function"""":""""github.com/evergreen-ci/evergreen/command.dirExists"""",""""file"""":""""/data/mci/8a4e334fe1089b7dbac8d2778c8083f8/gopath/src/github.com/evergreen-ci/evergreen/command/util.go"""",""""line"""":16},{""""function"""":""""github.com/evergreen-ci/evergreen/command.createEnclosingDirectoryIfNeeded"""",""""file"""":""""/data/mci/8a4e334fe1089b7dbac8d2778c8083f8/gopath/src/github.com/evergreen-ci/evergreen/command/util.go"""",""""line"""":26},{""""function"""":""""github.com/evergreen-ci/evergreen/command.(*s3get).Execute"""",""""file"""":""""/data/mci/8a4e334fe1089b7dbac8d2778c8083f8/gopath/src/github.com/evergreen-ci/evergreen/command/s3_get.go"""",""""line"""":155},{""""function"""":""""github.com/evergreen-ci/evergreen/agent.(*Agent).runCommandSet.func2"""",""""file"""":""""/data/mci/8a4e334fe1089b7dbac8d2778c8083f8/gopath/src/github.com/evergreen-ci/evergreen/agent/command.go"""",""""line"""":118},{""""function"""":""""runtime.goexit"""",""""file"""":""""/opt/golang/go1.13/src/runtime/asm_amd64.s"""",""""line"""":1357}]} ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7711","04/02/2020 14:43:06",1,"Occasional InvalidParameterValue errors from AWS API for AttachVolume ""Sometimes Evergreen gets back InvalidParameterValue back from the AWS API for an AttachVolume request https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Devergreen%20client%3D%22%5C*cloud.awsClientImpl%22%20message%3D%22AWS%20API%20call%22%20metadata.level%3D70%20%20api_name%3DAttachVolume&earliest=1585832400&latest=1585834200&display.page.search.mode=verbose&dispatch.sample_ratio=1&sid=1585834826.68129. At a minimum, Evergreen should avoid doing retries in this case. Also, is there validation that Evergreen can perform to avoid making this call in the first place?""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7715","04/02/2020 17:30:04",1,"Show spawnhosts' display name on termination ""EVG-7606 added display names to spawnhosts. We should use it in more places on the spawnhost page. For example, the terminate host modal could use it instead of the instance id. ""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7716","04/02/2020 18:12:19",1,"Spawn host user config file should be written as yaml instead of json ""This is mostly because a virtual workstation user (or any spawn host user, really) wouldn't be able to read the file. I think this would also cause some problems if the evergreen binary updates the file, which it assumes is yaml and not json.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9196","04/02/2020 18:49:54",2,"Configure selected tests pytest to run all the correct tests and checks ""As a DAG engineer, I should be able to run all the required pytests and checks in one command, So that selected-test development is stream lined. ---- AC: We already support flake8 and isort * configure pytest to allow it to run: ** black ** pydocstyle src ** mypy src * Update docs/Contributing.md to reflect the simplified command. * we will not be fixing the errors in the test directory""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8815","04/02/2020 18:52:22",8,"Allow processes to be created within containers ""We should be able to run processes in within docker containers. It should also manage shutting them down. We should have a new docker manager to handle these processes.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7719","04/03/2020 15:20:24",2,"Ask for a different AZ when AWS gives an InsufficientInstanceCapacity error ""EVG-7698 stopped retries for an InsufficientInstanceCapacity error. We still can't get hosts of the distro until that AZ gets capacity and it will still eat into the API limit because the host allocator still needs the hosts. As an alternative we could update the distro document cached in the host intent to use a different subnet/AZ on the next retry.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7722","04/03/2020 19:57:16",0,"Evergreen binary should be owned by distro user ""In user data, it might be curled by root and not the distro user so it can't be overwritten if the agent monitor needs to be redeployed.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7724","04/06/2020 15:37:14",1,"Writing .evergreen.yml file in home directory requires root permissions ""There was a change to host provisioning that prevents non-userdata spawn hosts from provisioning due to attempting to write the home directory's .evergreen.yml instead of putting it in \{{cli_bin}}. This is likely because the buildhost configuration AMI writes this file as root.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7725","04/06/2020 15:50:59",2,"Add volume display names ""There should be a way to name and manage volumes so you know what you're attaching to a host. Maybe another table in the UI?""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7730","04/06/2020 17:14:38",1,"IDE's shell doesn't source .bash_profile ""The integrated shell in vscode isn't a login shell so it doesn't source .bash_profile. One of the issues this causes is the evergreen binary isn't on the path. We should start the shell as a login shell: https://code.visualstudio.com/docs/editor/integrated-terminal#_shell-arguments""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7733","04/06/2020 18:31:09",2,"make mount point and directory configurable ""people should be able to configure where symlinks/userdata are setup""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7734","04/06/2020 18:37:20",0,"workstations should be easy to configure and start ""there are a couple of operations that should be built into the workflow: - if you start a workstation distro, then it should automatically have the """"workstation flag selected."""" - if you do not have an unexpireable host configured, then you should automatically be enabled as an unexpireable host - when the host has entered running state there should be a notification sent. ""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7737","04/06/2020 20:58:35",2,"Task count number should reflect number of results for query ""the number on the right, aka the total, currently reflects total of all tasks for the patch. It should reflect total patches for the specific query. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7738","04/06/2020 22:30:19",1,"collect more metadata in volumes collection ""created, deleted, expired, in use/not""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7748","04/07/2020 16:19:02",2,"Number of command calls should apply to pre/post blocks ""Some commands like host.create and generate.tasks limit the number of times you can call them. These limits currently only apply to tasks but not for pre/post, as well as task group setup/teardown/tasks (which should probably be checked separately from tasks). We should check the limits on those as well.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7749","04/07/2020 16:50:43",1,"Project validation should give warnings for dependencies refer to nonexistent buildvariants ""Commands and dependencies allow you to specify which build variants the task should run on or which dependency it is waiting for. There should be validation that such buildvariants actually exist and throw warnings if they don't.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7750","04/07/2020 18:39:36",2,"Remove requires field from project config ""The only project that used requires was the server and it's been removed at this point, so we should get rid of it since it's not a very useful feature. We should also remove from the wiki.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7751","04/07/2020 18:44:24",1,"Build variant and task names should not be allowed to have commas, or be called ""*"" or ""all"" ""We currently only ban bars (|) in the names of build variants and tasks. would be ambiguous in dependencies which use {{*}}, which means """"all tasks"""" or """"all build variants"""". We should also warn for commas or """"all"""" in tasks and build variants, because they would be ambiguous in the CLI invoked through a shell (because we allow users to specify comma-separated lists of tasks and build variants and """"all"""" is a magic keyword for running all variants or all tasks).""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7752","04/07/2020 18:50:53",1,"Task groups should not be allowed to be defined multiple times ""We have validation to ensure that the same name is not used for [build variants|https://github.com/evergreen-ci/evergreen/blob/df7927071e3cfd5381e076243d079037e3788437/validator/project_validator.go#L609-L616] or [tasks|https://github.com/evergreen-ci/evergreen/blob/df7927071e3cfd5381e076243d079037e3788437/validator/project_validator.go#L880-L897]. However, we don't do this same name uniqueness validation for task groups.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8521","04/07/2020 20:26:32",1,"retried exceptions should throw original exception ""When we retry endpoints, if we continue to get an exception, we should rethrow that exception instead of throwing a RetryException. The [tenacity docs|https://tenacity.readthedocs.io/en/latest/#error-handling] explain how to do this.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7758","04/09/2020 16:33:08",2,"CheckYamlStrict complains about patch_optional and status if used in dependency ""Example project config: If I run {{evergreen validate}} on this config, it gives this warning during {{CheckYamlStrict}}: WARNING: yaml: unmarshal errors: line 24: field patch_optional not found in struct model.copyType line 24: field status not found in struct model.copyType It gives the same error if you specify {{depends_on}} in the task definition instead of the buildvariant task list. However, if you call {{LoadProjectInto}}, the actual {{model.Project}} result has {{PatchOptional}} and {{Status}} properly populated for that dependency."""," tasks: - name: taskA commands: - command: shell.exec params: script: | echo hello - name: taskB commands: - command: shell.exec params: script: | echo goodbye buildvariants: - name: some_buildvariant run_on: - ubuntu1604-small tasks: - name: taskA - name: taskB depends_on: - name: taskA patch_optional: true status: """"*"""" ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7762","04/09/2020 18:44:49",1,"Ensure that tasks are not both patchable and patch_only ""Currently it's possible to specify a task is both patchable and patch_only for the same task (in essence, never run). We should give warnings to not allow this combination.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7765","04/09/2020 20:07:52",1,"Validate that func and command are mutually exclusive in a command list ""If you try to validate a command that has both {{command}} defined and {{func}} defined, it won't error but the agent will ignore the {{command}} defined in favor of {{func}}. We should add warnings to the validator that the two options are mutually exclusive for a single item in the commands list.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7769","04/10/2020 15:52:51",2,"Multiple events created for TASK_FINISHED event ""https://mongodb.splunkcloud.com/en-US/app/search/search?earliest=1585195200&latest=1585281600&q=search%20index%3Devergreen%20message%20%3D%20%22processing%20event%22%20event_type%20%3D%20TASK_FINISHED%20event_resource%20%3D%20mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_inmem_display_parallel_38e017530f4a317c109524bde9bf1f1366896bc5_20_03_26_15_13_52&display.page.search.mode=smart&dispatch.sample_ratio=1&display.general.type=events&display.page.search.tab=events&sid=1586530279.467449 This led to duplicate tickets https://jira.mongodb.org/browse/BFG-575333 and https://jira.mongodb.org/browse/BFG-575334.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7772","04/10/2020 18:00:14",1,"Render errors on patch page ""Render full page error if problem loading patch. Render banner at top of page for all other errors ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7773","04/10/2020 18:03:32",1,"Error, success, warning and info banner component and context ""Pages: patch, task Components: patch tasks, patch changes, task logs, task tests, task files To do: if there is an error, show error banner in place of component""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7774","04/10/2020 18:38:03",1,"If user specifies volume, should spawn host in that zone ""Otherwise it'll fail to attach the volume""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7776","04/10/2020 19:09:09",1,"PatchMetadata resolver on taskResolver throws error for base commit ""FindPatchById method only accepts mongo ObjectIds. This error will happen on all task base commits because the base commit is a version and versions do not have ObjectIds. Solution: use FindVersionById method in PatchMetadata resolver instead of FindPatchById""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7777","04/10/2020 20:12:34",2,"Add drag-and-drop to project commands ""Splitting from EVG-7732 because it's more annoying to do than I anticipated. Commands should be rearrangeable because order matters, so adding a command is super annoying. Tried [https://material.angular.io/cdk/drag-drop/overview] but there were bugs adding it.. Found [https://www.npmjs.com/package/ng-drag-drop] for a new attempt but use whatever!""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8034","04/11/2020 00:39:33",1,"Host provisioning jobs with scopes cause job backups if unable to finish successfully ""I created a spawn host with a volume that doesn't exist in AWS since it was available in the spawn host UI (it's probably a bug to allow people to try attaching volumes that don't exist in AWS) and it got stuck in provisioning (couldn't finish the user data done job because {{attachVolume}} would always fail on the AWS call). This seems to have caused amboy to refuse to run other jobs in the queue. Splunk errors: https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Devergreen%20i-0b3a0f488d280099a&display.page.search.mode=smart&dispatch.sample_ratio=1&earliest=1586550600&latest=1586561220&sid=1586561901.485818""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9015","04/11/2020 00:50:52",1,"add logging on dispatch errors ""add something here: https://github.com/mongodb/amboy/blob/master/queue/driver_mongo.go#L791 possibly removing, https://github.com/mongodb/amboy/blob/master/queue/driver_mongo.go#L525 and better annotate: - https://github.com/mongodb/amboy/blob/master/queue/driver_mongo.go#L493 - https://github.com/mongodb/amboy/blob/master/queue/driver_mongo.go#L504 ""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7784","04/11/2020 01:02:38",1,"Host should be decommissioned if volume does not exist in AWS ""This is to avoid the problem where the user data done job keeps retrying but can never succeed since there's no such volume in AWS. AWS will return """"InvalidVolume.NotFound"""" for this issue.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7785","04/12/2020 16:14:23",1,"add admin configuration for setting the priority option on amboy queue ""UI/rest/etc.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7787","04/13/2020 15:00:39",1,"Investigate volume not found in DescribeVolumes ""[https://mongodb.splunkcloud.com/en-US/app/search/search?earliest=-15m&latest=now&q=search%20index%3Devergreen%20InvalidVolume.NotFound%20NOT%20api_name%20%3D%20DescribeVolumes%20DescribeVolumes%20%7C%20stats%20count%20by%20host_id&display.page.search.mode=smart&dispatch.sample_ratio=1&display.page.search.tab=statistics&display.general.type=statistics&sid=1586784655.609498]""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8355","04/13/2020 16:20:21",2,"Signal processing algorithms deploy failing ""ex: https://evergreen.mongodb.com/version/signal_processing_algorithms_75312e873543f0f89aace75f43ded783395425c5 As a DAG engineer, I'd like poetry deploy to succeed for mainline builds, such that, we don't see red in evergreen, and the binary wheels upload successfully. AC: * Deploy does not fail in mainline""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7792","04/13/2020 19:32:49",1,"Delete distro teardown scripts ""Distro teardown scripts are no longer used by any projects. We should delete the code supporting distro teardown.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7797","04/13/2020 22:13:50",1,"Stopping host should detect when instance is stopped without depending on timing ""[This piece of code|https://github.com/evergreen-ci/evergreen/blob/f8d313109ff81e1413df35455dad402a7aa674a2/cloud/ec2.go#L944-L975] attempts to stop the host but does not do so in an idempotent way. If the check for host status fails (which is not guaranteed to succeed since it depends on how long the host takes to shut down), the host will be stuck in the stopping state with no way to recover. We should check if the host is stopped every time this is called and not just blindly make an AWS call which may fail. We also should unset the stopping state and revert to the old state if it fails. This code should not rely on sleeping as a reliabl mechanism to try to give the host time to shut down. ""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7798","04/13/2020 22:27:20",0,"Check in .prettierrc ""check in prettierrc so our editors format spruce code the same way""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7800","04/14/2020 14:43:55",1,"Communicator should print the body text verbatim if it is not a gimlet.ErrorResponse ""Many of the communicator routes attempt to parse a {{gimlet.ErrorResponse}} if a REST request returns something other than 200 OK. However, it's not guaranteed that the app server will even return a body in that format, and in those cases, the {{Communicator}} oftentimes returns unhelpful messages about parsing errors from gimlet, most commonly {{""""invalid character 'x' looking for beginning of value""""}}). In that case, we should always fall back to returning the response body as the error text if a gimlet response can't be parsed.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9211","04/14/2020 14:57:06",1,"Shrub should support task sync commands ""This should implement s3.push and s3.pull.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7802","04/14/2020 14:59:36",1,"Expose commit date for a version in expanded metrics triage UI ""As a DAG engineer, I'd like to know the date of a version's commit in the expanded metrics triaging UI, such that, the chronological sorting being done is obvious. AC: * version's commit date exposed""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7803","04/14/2020 15:14:34",2,"Add variant/task/test links to expanded metrics triaging ui ""As a DAG engineer, I'd like there to be links for each change point entry on their variant/task/tests, such that, you can easily navigate to those pages without having to do URL surgery. AC: * for each change point, variant/task/test are linked (see legacy UI for example)""","",0,1,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7804","04/14/2020 15:36:30",0,"Remove unused catcher in MergeGeneratedProjects ""A catcher catches errors but doesn't do anything with them.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8450","04/14/2020 18:13:39",1,"Shrub.py uses the wrong key for setting exec_timeout_secs ""Shrub.py allows you to [set exec_timeout|https://github.com/evergreen-ci/shrub.py/blob/05c234280f31bd5312b7d48e047ab339c7f34873/src/shrub/config.py#L159], but when it represents the value, it [stores it|https://github.com/evergreen-ci/shrub.py/blob/05c234280f31bd5312b7d48e047ab339c7f34873/src/shrub/config.py#L48] in the """"timeout"""" key which I understand to not have this meaning.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7805","04/14/2020 18:39:51",1,"Don't alert on starting spawn host ""Evergreen sends 2 alerts, one on starting a spawn host, and one on it successfully provisioning. A user doesn't care about the first alert, because there is nothing they can do yet. We should remove the first alert.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7809","04/14/2020 21:49:24",1,"Evaluate idle host termination for user data hosts stuck in provisioning that do not have work ""This host (i-0e069286dc0256011) got stuck in provisioning due to an app server restart when it was trying to requeue the job to retry provisioning. The idle host aggregation assumes that hosts are transiently in provisioning and will either fail to provision (decommissioned) or succeed (running). It doesn't handle a host which is permanently stuck in provisioning but has an agent running (because the LCT check will always filter out a host that's running an agent), so we should let user data hosts be considered for idle termination if the agent is running and not running tasks.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8383","04/14/2020 22:14:24",1,"Poplar runs out of memory when uploading large files ""In [this task|https://evergreen.mongodb.com/task/sys_perf_linux_standalone_insert_remove_patch_c4bca20e65afbb77834559247bdda8f541cda16b_5e93542f7742ae582b499d67_20_04_12_17_49_57##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%2522c4bca20e65afbb77834559247bdda8f541cda16b%2522%257D%255D%257D] we called {{poplar send}} with the results of a 12-hour run of InsertRemove and ended up running out of memory.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9151","04/15/2020 16:48:40",1,"set better batch size for amboy queries in mongodb ""it's currently 100, and when there are scopes this is probably wrong? maybe we should just let it be default? ""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7810","04/15/2020 18:01:09",2,"Patch: Reconfigure patch UI ""Designs: [https://mongodb.invisionapp.com/d/main#/console/19647524/411104420/preview] UI for selecting build variants and tasks to configure a pach""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7811","04/15/2020 18:02:23",1,"Patch: Schedule patch UI ""Clicking on """"Schedule"""" button in the patch reconfigure page schedules the patch and redirects user to the patch page of the newly scheduled patch""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7812","04/15/2020 19:41:59",2,"Add description, author, enqueue time, and code changes to commit queue query ""Most can be found on the patch data structure, but the enqueue time is an additional field on the commit queue item that needs to be exposed""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7813","04/15/2020 19:52:22",2,"Run user data on restart for Windows without relying on handleUserData being reset ""We're currently relying on a script in buildhost-configuration to reset the user data flag indicating whether user data has run yet in order. However, the Windows instances will no long be resetting the user data to allow evergreen to run user data again. We should add the persist tag to the end of the script to ensure it runs when it reboots. I think sommanye of the operations have not been built with the intent to be run multiple times (e.g. setup script logic, echoing to authorized keys for spawn hosts), so we should make sure that these operations are resilient to running multiple times.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7814","04/15/2020 21:12:20",1,"Distro page gives error if you re-create a distro that previously existed ""If you delete a distro, then recreate a distro of the same name, there's a flash that says """"error adding distro"""", which seems to be caused by [these lines|https://github.com/evergreen-ci/evergreen/blob/4d8f4df42b6f60057737b9d3a8ccee3a56815d72/service/distros.go#L331-L339] Logs: https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Devergreen-staging%20%22error%20adding%20distro%22&display.page.search.mode=smart&dispatch.sample_ratio=1&earliest=-15m&latest=now&sid=1586981483.723360""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7817","04/16/2020 14:09:27",2,"Handle task sync with specific status ""Right now, the task will sync if it's requested from the CLI regardless of whether it succeeded, failed, system-failed, or setup failed. We should add an option to specify under which conditions it should sync. {{ShouldSync}} should probably be refactored as a struct of options for task sync for a patch.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7819","04/16/2020 14:43:53",2,"ensure restarted tasks unblock blocked dependencies ""we have observed that some tasks become blocked by the dispatcher after their dependencies fail, and should become unblocked after one of their upstream dependencies become unblocked. (and subject to the same, potentially eventual, reblocking if they're still unsatisfiable for other reasons (age, other blocked dependencies.)) ""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7822","04/16/2020 16:59:50",2,"Correct percentage/ratio calculation in perf build baron view ""This value should be calculating the %change of the magnitude of the change, either a drop or an increase. This [line|https://github.com/evergreen-ci/evergreen/blob/3fc333d4f9cda59c22bf156cb86efc10e0196d69/public/static/app/perf-bb/hazardLevelCell.js#L10] should just be AC: * Need to change the column heading to %change from hazard level * Add a test ""","return 100 * ((a/b) - 1) ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8848","04/16/2020 17:09:45",1,"support transition to appropriate REST methods for authentication workflow ""Cedar historically supported GET http verbs for authentication workflows despite the fact that the data payload for these requests were in the body of the request. This error was corrected in MAKE-1173, and we inventoried all known clients, but have missed some externally maintained clients. To resolve, we should (quckly) add back support for GET verbs (in addition to current support for POST) in these cases, and then when the old clients have been fixed and sufficently back ported we can remove them. I would prefer to leave this ticket open to track removing support for GET in a few days/weeks ""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7911","04/16/2020 19:32:46",2,"can't spawn a rhel70-cloud from failing task ""https://evergreen.mongodb.com/task/mms_code_health_COMPILE_BAZEL_patch_bcac666f557161e095706b0ff113d2c28df4ee41_5e989c493627e01ffa20566f_20_04_16_17_56_27""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7827","04/17/2020 16:09:23",0,"virtual workstation setup commands should be scoped to the distro's home volume ""also prefix, when possible (maybe optionally), by the project directory.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7828","04/17/2020 18:29:24",2,"Let user specify max timeout on task sync from CLI ""If requested from the CLI, task sync is given an hour to finish. If the one hour timeout isn't generous enough, they can set it higher.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7829","04/17/2020 19:21:46",2,"Allow distro user data to specify type of user data ""If none is specified, it won't do anything. However, if you specify the user data type, it can do checks on whether or not the correct tags and directives are included. This is mostly to solve an incompatibility between host.create (which requires that we _not_ provision the host with user data) and the distro user data (which has to be merged with the type of the existing user data, but can't do so if there is none).""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7831","04/17/2020 19:49:53",1,"Virtual workstation never expire checkbox is between volume settings ""Maybe it could go under the volume size?""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7832","04/17/2020 21:07:48",2,"Virtual workstations are killed shortly after provisioning ""[This commit|https://github.com/evergreen-ci/evergreen/commit/5da06af4cd6f1ea0483968fd08878ff7e4531c11] puts a host termination job on the queue when provisioning fails. When copies of the job run concurrently the first job to successfully mount the volume will succeed and mark the host as running, but later jobs will try to mount an already mounted volume and fail (because the script isn't idempotent). The later job puts a job on the queue to terminate the host and it's killed a few minutes after it's been marked as running.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7835","04/17/2020 23:37:25",2,"app shouldn't fail to start if an api_only user is misconfigured. ""alert, and just don't allow that user to log in, but oof failure to start for bad data is heart wrenching. ""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8645","04/19/2020 19:10:06",2,"improve debugability of chains of Command executions ""when you get an error and the command has many subcommands it's difficult to know what ran""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8550","04/19/2020 19:12:43",2,"add conditional ""When"" methods for adding arguments to reduce complexity of calling code ""a la LogWhen in grip, all of the methods that add commands should have conditional options""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8524","04/20/2020 14:49:01",1,"add logging if Next takes longer than a second ""number of jobs considered, time, job type dispatched.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7836","04/20/2020 15:10:34",1,"Proof of concept: legacy/new UI toggle ""Write proof of concept for toggling between new UI and legacy UI Go into codebase and assert that proposed design for toggle will work""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7837","04/20/2020 15:10:54",1,"Proof of concept: New Relic analytics ""Write proof of concept for how analytics will be utilized What questions are we trying to answer through analytics? How will analytics be implemented in the app? ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7838","04/20/2020 15:15:43",1,"Patch tasks table should have initial sort value ""Problem: clicking patch tasks table sorter does nothing on the first click because table sorter changes from default to ASC. This is a jarring experience because nothing occurs as result of user action. Solution: task tests table has an initial state for sorters. Implement same initial state logic for patch tasks table as seen in task test table. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7839","04/20/2020 15:16:43",0,"Remove project dropdown from navbar ""Project dropdown is used for waterfall page. Until waterfall page exists, remove dropdown from navbar""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7840","04/20/2020 15:17:13",1,"Show Evergreen icon and logo in navbar ""Clicking on logo or icon goes to 'my patches' page""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7842","04/20/2020 15:17:39",1,"Favicon ""app should have favicon that matches evergreen logo""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7843","04/20/2020 15:24:57",2,"""Cannot redefine tasks"" error in generate.tasks ""https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_sharding_ese_gcm_gen_patch_a94df7c8c6f36e67086406b5778368f97dc1bd7f_5e9da7109ccd4e313742fcb6_20_04_20_13_44_02/0##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%2522a94df7c8c6f36e67086406b5778368f97dc1bd7f%2522%257D%255D%257D""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7844","04/20/2020 15:39:13",1,"Proof of concept: release strategy ""When will new UI be released? How will new UI be released? To whom? What comes after that What needs to be updated for the release to be possible? Who needs to know about the release? How will the release be communicated? Who will do the communicating for the release of the new ui? Map out the entire flow, from beginning to end, of the features included in the release. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7848","04/20/2020 17:07:47",2,"User data does not appear to run on ubuntu arm64 hosts anymore ""Recently (i.e. within the past week or two), user data has stopped running on ubuntu 1604/1804 arm64 hosts. I had to switch these hosts to use new SSH provisioning for now, but we should figure out why the user data scripts aren't running anymore. They used to be able to provision with user data.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7851","04/20/2020 17:35:41",2,"Nav: dropdown ""dropdown showing all links for user - link to legacy ui for all links unsupported in new ui (which for first release will be all of them) ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7853","04/20/2020 17:47:30",2,"Tables: statuses should autoupdate ""statuses in these tables should autoupdate""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7855","04/20/2020 18:27:26",2,"Add validation to generate tasks that variants must have a display name ""The approach that suggests itself to me is that the project validator should validate that variants have a display name, since generate.tasks calls the validator. We should make sure that this doesn't break any existing projects.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8900","04/20/2020 23:34:34",2,"Return iterator when downloading test results ""This way if there are thousands of results, we are not loading them all into memory""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7859","04/21/2020 14:41:12",1,"Add ""First Task Failure"" notification to PR patch runs ""I noticed the addition of """"First Task Failure"""" notifications on the evergreen user settings. This would be great if the functionality can be added to add to PR runs also This would be very useful for some of us – the complete PR suite for mms can take about 2 hours, and the first failure may happen in 10 minutes. This would easily allow us to get early notice if a PR has caused any failures without having to keep checking on the PR page.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7861","04/21/2020 15:27:41",2,"Request context canceling for logs ""With really long buildlogger logs, the request context cancels before the logs can fully load. We should fix this.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7865","04/21/2020 22:05:35",1,"The EVG cli is complaining about invalid fields (status) ""h3. Repro 1. Run the following commands: h3. Context It looks like the status field should exist: https://github.com/evergreen-ci/evergreen/blob/e41b1bdef0f6b0ca8140abe0885528a0333764c3/rest/model/task.go#L39 (it also appear in the EVG wiki) and in this test https://github.com/evergreen-ci/evergreen/blob/e7afea271efe1673e5f6f16fa83f38cbd436f526/model/project_parser_test.go#L55 """," git clone https://github.com/10gen/mms cd mms # Reproducible at this SHA, using the current evg cli (2020-04-15) git checkout c27242a8800c7a560576d4b43e616480cdcb9496 #Delete this line https://github.com/10gen/mms/blob/d41fcb4fac5534d2b0146d7bbc91586432a41033/.evergreen.yml#L272 # Validate evergreen validate ../.evergreen.yml WARNING: yaml: unmarshal errors: line 2866: field status not found in struct model.copyType ... ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7871","04/22/2020 19:10:52",2,"Reconfigure Page: Used mixed state checkbox for select all and deselect all ""select all and deselect all is currently controlled by a button. it should be a mixed state checkbox""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7874","04/22/2020 22:39:10",2,"filter/report recent task statuses by agentVersion ""version and revision. ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7879","04/23/2020 22:04:04",2,"Prevent the removal of a home volume ""If a user removes their home volume the system won't boot (because the device is written into the /etc/fstab file)""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8814","04/23/2020 22:38:41",8,"Implement jasper job dependency manager ""This is to handle dependencies in jasper build.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8582","04/23/2020 23:03:52",1,"Allow evergreen logging backends to be configured for commands from build file ""Evergreen offers different logging back-ends. We should expose this in the build system config files so people can configure evergreen logging as necessary.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7880","04/24/2020 14:48:23",1,"Cypress: cy.login fails when tests rerun in dev mode ""the cypress test runner reruns tests in the test browser and preserves cookies when code changes are made during local development. Rerunning tests that use cy.login fail because the command is unable to find credential inputs since the logged in user is redirected away from the login screen. cy.login should be modified so it doesn't search for inputs if the relevant cookies exist""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7882","04/24/2020 15:56:13",1,"Set the commit when switching branches through API ""When using the API to switch branches, the commit should be set as well so the build can run successfully. context: https://mongodb.slack.com/archives/C0V896UV8/p1587738944466200""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7883","04/24/2020 15:59:29",2,"Many repetitions of post-task logs ""In [this|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_unittests_14a5bcab6ee67dad278c02d74fea2bf4df149aa2_20_04_08_23_24_39] task's logs there is a very large section that is repeated over and over with the same messages and same timestamps (for most of the log's page), including the line {{Finished running post-task commands}}.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7884","04/24/2020 16:38:16",0,"Pull from task sync to spawn host seems to miss files ""When requested, {{evergreen pull}} runs on the spawn host, but seemingly misses several files because the process exits with code 1. Jasper process info from a failed pull attempt: { """"id"""": """"ffd19d7c-68df-4cb0-b87b-a1bf8c4bf272"""", """"host"""": """""""", """"pid"""": 671, """"exit_code"""": 1, """"is_running"""": false, """"successful"""": false, """"complete"""": true, """"timeout"""": false, """"options"""": { """"args"""": [ """"/bin/bash"""", """"-l"""", """"-c"""", """"/root/evergreen -c /root/.evergreen.yml pull --task mci_lint_lint_model_patch_783fc9cfbe5d65fdd7c8c61827e0a8cd65fa4b44_5ea2eb1397b1d369722e887e_20_04_24_13_35_40 --dir /data/mci"""" ], """"env"""": { """"JASPER_ID"""": """"ffd19d7c-68df-4cb0-b87b-a1bf8c4bf272"""", """"JASPER_MANAGER"""": """"24c0a80c-eb17-495e-95a4-545c5ef7d2d6"""" }, """"synchronized"""": false, """"implementation"""": """""""", """"working_directory"""": """"/"""", """"output"""": { """"suppress_output"""": false, """"suppress_error"""": false, """"redirect_output_to_error"""": false, """"redirect_error_to_output"""": false }, """"stdin_bytes"""": null }, """"start_at"""": """"0001-01-01T00:00:00Z"""", """"end_at"""": """"0001-01-01T00:00:00Z"""" },""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8501","04/24/2020 16:42:12",1,"Set StartAt and EndAt for RPC ProcessInfo and do not panic RPC conversion methods ""These fields were defined for local ProcessInfo but are never set in the protobuf. We also should not panic during conversions, which I've just noticed in jasper.ext.go, because panics will cause the RPC service to return no error and no response (which is not helpful) when it abruptly shuts down. Instead, it should return some error during the RPC-Jasper model conversion. Relying on these methods on the client side will also cause panics.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8449","04/24/2020 16:47:06",1,"Get change points by version is failing/bugged ""Can't get a list of change points grouped by version from the cedar api. As a DAG engineer, I'd like to be able to get a list of change points grouped by version for any project, such that, I can do things with these lists. AC: * this route works for page 0 * page_size can't be 0 * page_size is handled properly""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7887","04/24/2020 18:07:52",2,"Log AWS retries due to request limits at lower log level ""Since retries due to request limits are an expected condition, we should not log these at error level.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7895","04/27/2020 13:42:47",0,"Pull should error if no matching task directory is found in s3 ""Right now, it'll just no-op and pull nothing. However, it should be an error to pull if the task sync contains nothing in it.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7896","04/27/2020 15:35:17",8,"Adjust patchTasks, userPatches and taskTests pagination to account for recently created/deleted documents ""Paginated infinite scrolling tables powered by the patchTasks, userPatches or taskTests queries can show duplicate rows if patches, tasks or tests are added or deleted from the db while scrolling. A reason for this is because we currently count the number of documents to skip in our paginated db query based on how many rows we are showing in the UI""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7897","04/27/2020 16:56:07",1,"remove ssh_timeout_secs from evergreen.Settings.HostInit ""it's unused. ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7899","04/28/2020 03:18:15",2,"Disable color shading in Perf Discovery ""When selecting expanded metrics in perf discovery the ratio column is still colored based on the assumption that higher is better. Numbers greater than 100 are green, and those below are red. For the latency values reported by expanded metrics, this is not correct. This ticket is to disable the coloring until such time that EVG-7697 is completed. Acceptance Criteria: - No more colors in perf discovery for any metrics. ""","",0,1,0,0,0,0,0,0,0,0,0,1,0,0,0
+"EVG-7904","04/28/2020 15:27:07",0,"Write better doc comments explaining meaning of host timing constants ""There are several time constants used in host management that have subtle differences (e.g. {{provisioningCutoff}} vs {{MaxLCT}} vs {{MaxUncommunicativeInterval}}). It would be nice to explain what they are in a doc comment so people know what the exact semantics are and don't accidentally misuse them.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7907","04/28/2020 18:15:28",1,"My Patches: Return total patches from userPatches query ""return the total number of patches without factoring in Page & Limit from the userPatches query""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7972","04/28/2020 18:19:25",1,"Spawnable selenium server image ""Create an Evergreen container distro that has this as its image [https://hub.docker.com/layers/selenium/standalone-chrome/4.0.0-alpha-5-20200409/images/sha256-8e6222259f4098dfc9440b0d57b04088c3ea750eb6628b973fac17a646ea11f0?context=explore.|https://hub.docker.com/layers/selenium/standalone-chrome/4.0.0-alpha-5-20200409/images/sha256-8e6222259f4098dfc9440b0d57b04088c3ea750eb6628b973fac17a646ea11f0?context=explore]""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9110","04/29/2020 16:18:19",2,"Jasper remote tests should test REST client and scripting/logging should actually be implemented ""The REST interface was never added to the remote manager tests for some reason. I tried to add the REST client to the remote tests but it fails most of the scripting and logging tests, which were untested and don't work according to the test cases.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7913","04/29/2020 18:00:53",2,"Mark tasks as failed if the build fails during ""go test"" ""When running {{go test}}, the build can fail. The Evergreen parser for {{go test}} output doesn't currently handle this, so none of the tests will be run and the task will succeed, even though it has technically failed. Example task: [https://evergreen.mongodb.com/task/mongo_go_driver_tests_42_plus_zlib_zstd_support__version~latest_os_ssl_32~ubuntu1604_64_go_1_12_test_replicaset_auth_ssl_patch_a484eb6d9594fce7c5d6a15960a15cd04c03f986_5ea9aad10ae6062860bc3011_20_04_29_16_26_58##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%2522a484eb6d9594fce7c5d6a15960a15cd04c03f986%2522%257D%255D%257D] Relevant log line: [https://evergreen.mongodb.com/test_log/5ea9ad080305b924ba8232f5#L8676]""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7914","04/30/2020 03:05:09",2,"add flag to host or distro document to identify cluster members ""Evergreen has some distros which maintain hosts that don't run tasks and aren't container parents (the compile cluster for the workstations), but there is nothing in the data model that would track this, and allow us to report on it appropriately. The host idle data report should also reflect this ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8652","04/30/2020 16:45:10",2,"Upgrade cedar's test environment to mdb 4.2 ""As a DAG engineer, I'd like to use MDB 4.2 in cedar's tests, such that, we can leverage the features of 4.2, and such that the test environment matches the production environment. AC: * move from 4.0->4.2 in cedar's tests""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7918","04/30/2020 16:45:29",2,"Blacklist the commit queue's merge task when a task fails ""If an item is manually removed by a user we [blacklist the merge task|https://github.com/evergreen-ci/evergreen/blob/d79e6eb3d7e3e1565efbc5d9188ca4dad3e7a836/model/commitqueue/commit_queue.go#L257] so the item won't be merged if one of the tasks is restarted. We should also do this when we [abort the version for a failing task|https://github.com/evergreen-ci/evergreen/blob/d79e6eb3d7e3e1565efbc5d9188ca4dad3e7a836/model/task_lifecycle.go#L561].""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8403","04/30/2020 19:01:28",2,"Fix redirects in cedar ""I POST to cedar.mongodb.com/rest/v1/admin/perf/change_points and get redirected to GET cedar.mongodb.com/rest/v1/admin/perf/change_points. This seems to happen for other routes as well. I'm guessing the issue is auth? If so, unauthenticated requests should get a 401, not a 301 to a GET route that doesn't exist.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8980","04/30/2020 19:03:57",2,"Benchmark existing pail implementations ""This would be useful to determine what workloads the current implementations are best at handling. Most interesting would be the performance of uploading a large number of small files or a small number of large files. This has implications for how performant task sync is at uploading/downloading entire directories.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7921","04/30/2020 20:21:20",1,"Stop looping on unprocessed notifications ""[This commit|https://github.com/evergreen-ci/evergreen/commit/0290a7087d74047154ac8abbef6201f5de27e07d] tries to dispatch notifications that were dropped by their event-send job (if the job's context is cancelled, for instance). However, there's a collision in the job ids and the new job isn't inserted. The event-send job's id should include a timestamp.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7926","04/30/2020 23:00:40",2,"add more logging when we terminate a provision failure host ""we seem to only have logging when setup scripts run, and it's possible that these kinds of events are centrally logged in a meaningful way.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8694","05/01/2020 03:28:47",2,"put integration tests behind flag to help major ""put things that depend on external services like aws behind a flag to improve local development experience, so that you don't need credentials to run tests. open question: should we also have a flag to exclude tests that depend on a running mongodb instance or is that a reasonable assumption?""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7930","05/01/2020 14:52:28",1,"GetInstanceStatus spins if the host ID is no longer in AWS ""GetInstanceStatus should return terminated if it receives InvalidInstanceID.NotFound. If enough time has passed that the host is no longer in AWS, this means that the host has in fact been terminated, but currently GetInstanceStatus retries on this error.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7934","05/01/2020 18:02:17",2,"spread historic repotracker build better ""- reduce frequency of catchup jobs more. - activation should be less of a spike. and could happen in fewer jobs as they're pretty quick. ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8442","05/01/2020 19:17:49",2,"Refactor expanded metrics front+back end not to rely on thread levels directly ""As a DAG engineer, I'd like thread levels not to be a first class citizen in terms of the data models we use, such that, cedar doesn't need to worry about arguments to test runners individually. AC: * filtering by thread level is done in a way such that the API does not explicitly know about thread levels, but instead applies filters for arguments (e.g. arg_filters = {thread_levels: [1,2,3]} rather than thread_levels = [1,2,3]) * CPs returned on the version_with_change_points route have arguments on them, rather than thread levels * The us takes thread levels from the arguments on a cp, rather than directly from a cp""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7938","05/01/2020 22:35:07",1,"When spawning a host from a task, the host spawn dialogue reappears ""If you spawn a host from the """"Spawn..."""" link on a task page, after clicking Spawn from the modal, the host is spawned and then the modal reappears. I don't see how, but I suspect it was introduced by https://github.com/evergreen-ci/evergreen/commit/3b681729e27b6b09320ed1158f6c3eb39d75bf6d, since that's the only recent change I see that would affect this page.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8781","05/03/2020 16:41:10",1,"add more dispatching statics to job complete notes ""- job dispatch time (time.Started-time.Created) - total job pending time (time.Finsihed-time.Created)""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8868","05/04/2020 14:56:49",2,"Create Jasper build CLI that can generate evergreen configs from golang build files ""Mostly, this should just be feeding a yaml config file into the golang evergreen config generator. For now, this should just be a convenience to test the generator quickly. We should also create an explicit Jasper binary in {{cmd}}.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7945","05/04/2020 16:20:27",2,"Error reporting proof of concept ""figure out which errors should be reported to bugsnag and their severity level""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7946","05/04/2020 16:36:54",2,"Mount volumes by UUID ""Currently volumes are mounted by device name, which is subject to change (e.g if a user changes the instance type). Because this is written into the instance's /etc/fstab file changes to the device name prevent the instance from booting. Instead, the /etc/fstab should reference the volume's UUID. https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ebs-using-volumes.html#ebs-mount-after-reboot""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7948","05/04/2020 21:45:22",1,"Volume expiration query inefficient ""The query to find volumes that have expired does a lookup on an unindexed foreign field (volumes.volume_id)""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7950","05/04/2020 22:08:16",2,"tests and tasks tables should be paginated instead of infinite scroll ""also allow users to enter their own page size""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-9270","05/05/2020 16:46:27",0,"Server should be closed as soon as the context cancels ""Currently, it only checks for a done context before each connection accept so it can wait forever on accepting connections without cleaning up. We should have a thread that closes the server as soon as the context is done to clean up resources.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7954","05/05/2020 17:45:09",1,"Fix commit queue button sizing ""!https://files.slack.com/files-pri/T024FNNHU-F012WBPQ8BY/screen_shot_2020-05-05_at_12.40.41_pm.png! On a 1200 px width screen""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7955","05/05/2020 18:46:32",2,"lobster hides 'truncated' fields on log lines ""When a server log line is too long, Structured Logging will truncate the line by omitting any fields past the line limit, and instead adds two new fields at the root level of the JSON, """"truncated"""" and """"size"""". Lobster is hiding these fields, which makes it hard to tell what lines are actually truncated when viewing a log. Here is an example, search for """"T00:19:23.903"""" to compare the raw log output versus the HTML Lobster output: https://logkeeper.mongodb.org/build/bb9f3c027d7e502fb455308955e19b5a/test/5eaf5f88c2ab684da28bda30?raw=1 https://logkeeper.mongodb.org/lobster/build/bb9f3c027d7e502fb455308955e19b5a/test/5eaf5f88c2ab684da28bda30#bookmarks=0%2C29904&l=1""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7957","05/05/2020 20:53:51",1,"My Patches: prevent pagination component Next button from moving ""the next button moves from right to left when paginating over many pages. ""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-7960","05/06/2020 16:16:14",2,"Batch Fleet Requests ""The CreateFleet request takes a slice of LaunchTemplateConfigs. We should investigate if this allows us to batch CreateFleet requests. Uploading the launch templates (CreateLaunchTemplate) will probably still need to be individual.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7964","05/06/2020 20:00:49",1,"Volumes UI doesn't refresh when a host gets a volume in the background ""When a new volume is added as a virtual workstation's home volume, the new volume which is linked from the host isn't present on the volumes page until the page is reloaded.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7969","05/07/2020 13:48:40",1,"subprocess.scripting needs to support running tests in the harness ""This is so that we can generate projects that use evergreen configurations from scripting generators.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7974","05/07/2020 22:30:50",2,"Allow task selection of multiple build variants at once in new UI ""I am not currently able to hold shift to select multiple build variants so I can hit """"select all"""" and select all tasks amongst those build variants. Will this functionality, or something equivalent, be available? Update: this ticket added functionality to select multiple build variants with Ctrl+Click (Command+Click on mac). EVG-8176 tracks the support for Shift+Click.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7975","05/07/2020 22:34:47",1,"Build variants and tasks can be selected while in ""Changes"" tab for patch configuration ""I am able to click on the different variants in the “Select Build Variants and Tasks” pane while in the Changes tab during patch configuration. Since the changes stay the same regardless of which build variant/task is selected, selecting a different build variant/task does not have an effect.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-7977","05/07/2020 23:27:35",2,"create Query for aws regions for user profile in spruce ""[https://github.com/evergreen-ci/evergreen/blob/master/service/user.go#L200] https://github.com/evergreen-ci/spruce/pull/162#discussion_r421199391""","",1,0,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8037","05/08/2020 16:00:11",3,"Virtual workstation home directories should be on the correct volume ""Virtual workstations currently have the users home directory at /home/ubuntu, but this is the small local volume. We want users to actually use the large volume. We should arrange things so that /home is mounted on the large volume, not the small one. Otherwise, it is too easy for tools that expect to write to things under $HOME to do the wrong thing. It is also unexpected for users that their $HOME directory is the wrong place to do work. ""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9228","05/08/2020 17:02:03",2,"Packages in server requirements should work with python 3.6 ""The server build system sets a python minimum of python 3.6, not python 3.7. However, shrub requires 3.7. This creates an unpleasant situation on Ubuntu 18.04 (which is the image type used for virtual workstations), where the system python3 is not acceptable for installing the `dev-requirements`. You can work around by using the toolchain python, which is py37, but this isn't great. We should ensure that our tools work on the most commonly deployed Ubuntu LTS without needing the mongodbtoolchain.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-7983","05/09/2020 01:58:11",2,"Host spawned from task missing artifacts ""When I spawn a host from this task, I don't seem to get any artefacts pre-loaded onto it: https://evergreen.mongodb.com/task/mongodb_mongo_v4.4_enterprise_debian10_64_logical_session_cache_replication_1sec_refresh_jscore_passthrough_3_enterprise_debian10_64_bbdf0a11d1c61be0760a829e82799129beac7be0_20_05_08_09_20_41/0 """," admin@ip-10-122-1-238:/data$ ls tmp admin@ip-10-122-1-238:/data$ ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9070","05/11/2020 18:22:39",1,"Deduplicate identical options for golang ""Since people can specify options as the package or variant level, we have to be able to overwrite existing options while preserving unspecified ones when merging variant options into package options.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9120","05/11/2020 18:31:31",2,"Allow tags to be applied to all packages by default and explicitly exclude packages ""All tags have to be explicitly listed for now, which requires defining the package path and the tag associated with it. We should also support tagging all packages with some set of global tags, then explicitly excluding particular packages from default tags.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8203","05/11/2020 20:44:17",1,"Move all projects to git hooks ""Some projects use the legacy repotracker. This means that commits show up much more slowly, and it prevents using more modern features. If it's possible to move all projects to git hooks, we should also open a ticket to remove the option to configure a project as repotracker-only entirely.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7985","05/11/2020 21:03:29",1,"Users with volume size >= their limit can't spawn any hosts ""If a user has volumes with a total size greater than the limit (currently 700GiB) they can't spawn hosts. The input [here|https://github.com/evergreen-ci/evergreen/blob/2127156a54b791ecd14e4a4d53101c7661294299/public/static/partials/user_host_options.html#L115] sets a minimum of 1. Because the input is disabled when the user doesn't have any remaining volume capacity the value is [stuck at 0|https://github.com/evergreen-ci/evergreen/blob/2127156a54b791ecd14e4a4d53101c7661294299/public/static/js/spawned_hosts.js#L702] and the form is invalid.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7988","05/11/2020 23:22:53",2,"Add helpful error message if evergreen commit queue merge fails due to GitHub branch protection ""From Slack: I enabled it and ran {{evergreen merge}} , but the commit queue merge test went faster than a different github PR check, so when evergreen attempted to merge it failed since not all required checks had completed. or at least this is what I assume happened. in github I only saw the message {{""""merge failed: """"}} ""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-7993","05/12/2020 19:44:44",2,"Unscheduling tasks/builds/versions in the commit queue doesn't remove the item for GitHub queues ""If a user aborts or unschedules a commit queue task we [remove the item from the queue|https://github.com/evergreen-ci/evergreen/blob/8b70262c283f271045c4d13f81ed98b0fa8e9680/service/task.go#L738]. We do the same [for builds|https://github.com/evergreen-ci/evergreen/blob/8b70262c283f271045c4d13f81ed98b0fa8e9680/service/build.go#L221] and [versions|https://github.com/evergreen-ci/evergreen/blob/8b70262c283f271045c4d13f81ed98b0fa8e9680/graphql/util.go#L434] The item identifier passed in is the version id, so it doesn't work for GitHub queues.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8636","05/13/2020 16:05:22",2,"periodically log statistics during event collection ""provide an implementation of the collector interface https://github.com/evergreen-ci/poplar/blob/master/vendor/github.com/mongodb/ftdc/events/collector.go#L20 so that they log aggregate stats (using grip) periodically, probably like """"running average of throughput and latency, and maybe some other elements that are in the events structure.) Considerations: - logging might impact the throughput of the event collector, which means we probably need to make it a wrapper to a collector. (This would probably be simplified by adding a Count() method to the Collector interface) - users may want to opt-in to this kind of reporting, which would require changes to the poplar registry and its interfaces. - for many tests, there are multiple collectors running in parallel, which means the logging messages produced should probably have an ID of some sort, again this might require adding a Name() method to the Collector interface, and providing a way for the to propagate through the interfaces. I suspect for clarity and stability, we will want to implement message.Composer (from grip) with a type that produces a well-formed message and optional . message.Fields may not have an easy to scan format. ""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8001","05/13/2020 16:56:16",2,"Evergreen's _archive.targz_pack_ command does not pick up all the files ""I am investigating a test failure: https://evergreen.mongodb.com/task/mms_e2e_opsman_os_sanity_E2E_OpsManager_SanityTests_RemoteAutomation_SUSE15_860d517f3684edeb52e65c76f75cc55d4e5c43f3_20_05_13_11_51_44 Our EVG config looks like this (mms/master): - command: archive.targz_pack params: target: """"logs.tgz"""" source_dir: """"mms"""" include: - ... - """"server/build/agents/logs/**"""" The generated logs.tgz file (on the task linked above) contains the following data: However, if you search all the Download logs, you will find this line: [junit] 2020-05-13T12:38:01.125+0000 [main] [com.xgen.svc.cukes.helpers.LocalAgentManagerFiles.INFO] [LocalAgentManagerFiles.java:47] - Renaming logsDir from server/build/agents/logs/42-Install-automation-and-backup to server/build/agents/logs/42-FAILED-Install-automation-and-backup *42-FAILED-Install-automation-and-backup is missing from the tgz* If you look at that tasks's history and download other logs.tgz files, you will find all kinds of inconsistencies (e.g. the previous one has two failed tests, but doesn't have a `server/build/agents/`directory. *This is seriously impacting our ability to debug why our test targets are failing!* Can you please look at this bug, with priority? Thanks, Mihai"""," ❯ ll build/agents/logs total 0 drwxr-xr-x@ 4 mihaibojin staff 128 May 13 16:42 41-PASSED-Install-automation-and-backup ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8685","05/13/2020 18:42:42",2,"simplify amboy marshaling flow ""use the thing julian and I have been working on for dependency injection in jasper""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8003","05/13/2020 22:51:24",2,"Add check box to variant select in patch restart modal ""!image-2020-05-13-17-51-00-199.png!""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8004","05/13/2020 23:11:01",1,"Use archive bucket in task sync ""This is to attempt to fix performance issues in task sync possibly due to uploading a large number of files.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8009","05/14/2020 18:15:37",1,"Use separate data set in evergreen for e2e tests and for developing locally ""Our e2e tests will make assertions based on a predefined set of data. Two separate data sets allows us to freely edit data for local development without worrying about breaking the e2e tests. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8010","05/14/2020 19:35:38",1,"Virtual-workstations are terminated shortly after provisioning completes ""Multiple user-data-done jobs running at the same time will sometimes cause the second job to terminate the host. When the job returns an error we [terminate the host|https://github.com/evergreen-ci/evergreen/blob/a2a5bb96591e9b2eecb09edb314fdf99abc53932/units/provisioning_user_data_done.go#L117]. attachVolume is mostly idempotent so a second job should just mount the volume a second time without erroring. Asking AWS to mount an already mounted volume is an exception because AWS will return an error if the volume is already mounted. We try to [continue on error for this case|https://github.com/evergreen-ci/evergreen/blob/a2a5bb96591e9b2eecb09edb314fdf99abc53932/units/provisioning_setup_host.go#L774], but it seems AWS sometimes returns a different error string: """"VolumeInUse"""". Previously we addressed this problem with scopes and now that scopes are fixed that may be a viable solution.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8011","05/14/2020 20:12:24",1,"Host start/stop/terminate buttons are misaligned ""it seems hiding uptime causes the buttons to move up a column. Why is uptime hidden, though? They're both running. (I just stopped/started the second one)""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8012","05/14/2020 20:15:52",1,"Update Host button adds an error flash if unexpirable hosts are at the limit ""I have two unexpirable hosts. The form isn't dirty, but the Update Hosts button is enabled. Clicking it displays an error flash that I can't have more than two unexpirable hosts. If I actually make a change I still get the error, but the change is made.""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8013","05/14/2020 20:19:50",1,"Volume create buttons are on the left of the modal ""Maybe they need to be float:right? Also, should the label be """"Create""""?""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8790","05/15/2020 14:59:24",3,"Make signal processing jobs in cedar report time series to Performance Analysis and Triage Service and remove all other tooling ""As a DAG engineer, I'd like to remove the signal processing stuff being done in cedar, and simply have it's [recalculate change points job|https://github.com/evergreen-ci/cedar/blob/master/units/recalculate_change_points.go#L69] migrated to only reporting time series updates to the Performance Analysis and Triage Service. The API for the reporting is as follows: The aggregation for getting this data [already exists|https://github.com/evergreen-ci/cedar/blob/master/units/recalculate_change_points.go#L89]. AC: * Cedar reports time series updates to performance analysis and triage service using the existing triggers for aforementioned job * All other signal processing tooling/schema removed * Remove change points from cedar's database Notes: * [Technical Design|https://docs.google.com/document/d/1wKLZ-IWjA8-kWAO44tHotzuYei0y2rNX7OQjrKGEo34] * [Design Modification|https://docs.google.com/document/d/1jLcNDgYPSJte11kfFnidWpFXfWUPnjy-zdq63UeWOQM]"""," POST $baseUrl/time_series/update { """"project"""" : """"sys-perf"""", """"version"""" : """"sys_perf_fb91e1d6e719de42905d63d26f065a4ac835be1e"""", """"variant"""" : """"linux-3-node-replSet"""", """"task"""" : """"parallel_insert_replica"""", """"test"""" : """"ParallelInsert-1024-Insert_WMajority_JFalse"""", """"measurement"""" : """"Latency99thPercentile"""", """"args"""" : [{ """"name"""": """"thread_level"""", """"value"""": 16 }], """"data"""": [ { cedar_perf_result_id: ObjectId(""""5eb31601bf3966ed0de32e04""""), order: 12311, value: 100 }, ... ] } ",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8378","05/15/2020 18:48:11",0,"Large writer closer should flush after it has reached the threshold size ""In [these line in the largeWriterCloser|https://github.com/evergreen-ci/pail/blob/a03af0d0a6184f872ec5ddd73631cb00f0ec3675/s3_bucket.go#L468-L474], we flush before we've reached {{maxSize}}. However, since this is the backing writer implementation of the multipart uploader, it should flush _after_ it has reached 5MB or else AWS will reject the multipart upload with the error {{EntityTooSmall: Your proposed upload is smaller than the minimum allowed size}}. We should change the semantics of the size threshold to mean {{minSize}} for uploading, rather than {{maxSize}} before we flush.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8027","05/18/2020 17:12:39",1,"mountLinuxVolume: umount command doesn't specify the absolute path to the device ""The device name no longer includes the absolute path. Fix the umount command to prepend the path (/dev/)""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8031","05/18/2020 21:00:30",1,"evergreen validate should error more helpfully ""Occasionally users have evergreen validate warnings from unmarshal strict that just look like """"error unmarshalling into temp structure."""" Would be nice if we could give them more context. [https://mongodb.slack.com/archives/C0V896UV8/p1589831494350700] ""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8035","05/19/2020 16:03:05",1,"Enable Git Tag Version Creation Checkbox on Project Page ""Similar to Enable Pull Request Testing, would be nice to have this as a checkbox so if someone is having an issue they can uncheck this rather than unpopulate related fields.""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8039","05/20/2020 02:01:59",1,"New provider settings don't log to the event log ""In https://evergreen.mongodb.com/event_log/distro/rhel76-docker, the AMI history does not show past April 6""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8041","05/20/2020 14:41:52",1,"Volume expiration warnings should link to the volume ""The link I got just took me to the spawn host page. It could link to the volume by setting the resource type and volume id parameters in the URL""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8043","05/20/2020 19:09:39",2,"Failed: coverage on Coverage [Evergreen @ 2c4eeec3] ""h2. [coverage failed on Coverage|https://evergreen.mongodb.com/task/evergreen_coverage_coverage_2c4eeec32698f70c87c4153f52a528ef58a4f1a5_20_05_20_14_24_51/0] Host: [ec2-52-86-97-72.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-06d9b27132960dd1f] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/2c4eeec32698f70c87c4153f52a528ef58a4f1a5]: EVG-7990 add git tag requester (#3602) | 20 May 20 14:24 UTC Evergreen Subscription: ; Evergreen Event: [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_coverage_coverage_2c4eeec32698f70c87c4153f52a528ef58a4f1a5_20_05_20_14_24_51/0?type=T] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-8044","05/20/2020 19:40:47",1,"Unexpirable hosts remaining message is confusing if it's broken over two lines """"""1 unexpirable host remaining"""" it's not clear the 1 is part of a message that got wrapped. Maybe we should set {code:css} white-space: nowrap; {code} [here|https://github.com/evergreen-ci/evergreen/blob/2c4eeec32698f70c87c4153f52a528ef58a4f1a5/public/static/partials/user_host_details.html#L99] and [here|https://github.com/evergreen-ci/evergreen/blob/2c4eeec32698f70c87c4153f52a528ef58a4f1a5/public/static/partials/user_volume_details.html#L43]"""," white-space: nowrap; ",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8045","05/20/2020 21:03:06",1,"Mounted volumes shouldn't display an expiration time ""... since they won't expire.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8529","05/21/2020 15:30:28",1,"Dist target should compile for Linux for deployment ""When you use {{make dist}}, it compiles but compiles the barque binary also based on your local GOOS (which is probably {{darwin}} for everyone except Sam). However, the barque binary in the {{dist.tar.gz}} needs to be built for Linux, so we have to run {{make build/make-tarball}}, followed by {{GOOS=linux make dist}} or else the binary will be built for macos. Also verify that custom local builds are deployable because it seems like the two most recent deploys for Amboy revendors did not actually restart the barque app server with the custom build (although they changed the {{/srv/barque/current}} symlink).""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8810","05/21/2020 16:49:11",1,"mciuploads/barque permissions issue ""https://slack-redir.net/link?url=https%3A%2F%2Fevergreen.mongodb.com%2Ftask%2Fbarque_ubuntu1604_dist_a7a85c58ecc01ff206816d3088acf74597348b67_20_05_21_15_44_13%2F0 looks like we should be able to upload, but maybe our key doens't have access""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9138","05/21/2020 19:14:47",0,"Move BuildFileList into utility ""This is so Jasper can reuse the function (will require libraries using utility to vendor {{github.com/sabhiram/go-git-ignore}}).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9235","05/21/2020 19:16:00",0,"Add yaml.UnmarshalStrict variant for utility operations ""The utility package provides convenience functions wrapping the regular {{yaml.Unmarshal}}, but we should also provide strict variants for stricter YAML parsing.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8052","05/21/2020 19:37:22",1,"Update Subscription Modal based on designs ""Update the Subscription modal based on designs from product designers""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8053","05/21/2020 20:51:09",1,"Aborted by task name is too long ""When a task is aborted on account of another task (for the commit queue) the task page includes a link to the task that really failed. The link text is the task ID which is unsightly and confusing. It should probably be """"_build display name_:_task display name_"""" or something. Example: [https://evergreen.mongodb.com/task/mongodb_mongo_v4.2_enterprise_rhel_62_64_bit_compile_patch_8dafa4060855b0568dbd3e103984d29cb2937628_5ec6d01c3627e07ed8130f62_20_05_21_19_02_01|https://evergreen.mongodb.com/task/mongodb_mongo_v4.2_enterprise_rhel_62_64_bit_compile_patch_8dafa4060855b0568dbd3e103984d29cb2937628_5ec6d01c3627e07ed8130f62_20_05_21_19_02_01##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%25228dafa4060855b0568dbd3e103984d29cb2937628%2522%257D%255D%257D]""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8690","05/22/2020 16:12:49",2,"jasper benchmarks fail to compile after logging configuration change """""," benchmarks/harness.go:89:51: undefined: options.Logger benchmarks/harness.go:91:26: undefined: options.Logger benchmarks/harness.go:96:14: undefined: options.LogType benchmarks/harness.go:97:16: undefined: options.Log benchmarks/harness.go:98:34: undefined: options.Logger benchmarks/harness.go:102:4: r.Begin undefined (type poplar.Recorder has no field or method Begin) benchmarks/harness.go:107:4: r.IncOps undefined (type poplar.Recorder has no field or method IncOps) benchmarks/harness.go:108:40: opts.Output.Loggers[0].Configure undefined (type *options.LoggerConfig has no field or method Configure) benchmarks/harness.go:114:4: r.End undefined (type poplar.Recorder has no field or method End) benchmarks/harness.go:122:15: undefined: options.LogType benchmarks/harness.go:122:15: too many errors ",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8056","05/26/2020 13:49:34",2,"Unable to run a local shell script in subprocess.exec on windows ""I’m trying to run a subprocess.exec in a windows host where the binary option is a local shell script but I’m getting an error Sample evg function {code:yaml}command: subprocess.exec type: test params: add_to_path: - ${go_bin} - """"/cygdrive/c/Program Files/go-msi"""" - """"/cygdrive/c/wixtools/bin"""" working_dir: src/github.com/mongodb/mongocli env: WORKDIR: ${workdir} CGO_ENABLED: """"0"""" <<: *go_options binary: build/package/generate-msi.sh {code} Task [here|https://evergreen.mongodb.com/build/mongocli_master_go_msi_snapshot_patch_4ed3c5b8e6ae45d996c033d688526965c0d3e7b8_5ecbde33d1fe0766215764e7_20_05_25_15_03_23] Logs [here|https://evergreen.mongodb.com/task_log_raw/mongocli_master_go_msi_snapshot_package_msi_patch_4ed3c5b8e6ae45d996c033d688526965c0d3e7b8_5ecbde33d1fe0766215764e7_20_05_25_15_03_23/0?type=T#L37] Is this possible for windows hosts? running shell scripts as a subprocess.exec ? I’m trying to move a lot of my shell.exec to be subprocess.exec this way but hitting this error with a windows host""","%1 is not a valid Win32 application command: subprocess.exec type: test params: add_to_path: - ${go_bin} - """"/cygdrive/c/Program Files/go-msi"""" - """"/cygdrive/c/wixtools/bin"""" working_dir: src/github.com/mongodb/mongocli env: WORKDIR: ${workdir} CGO_ENABLED: """"0"""" <<: *go_options binary: build/package/generate-msi.sh ",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8067","05/26/2020 19:49:48",2,"Track duration of time between patch build creation and patch build scheduling ""I am interested to track a metric that gives a sense of how long it takes a user to configure a patch build, whether via the evergreen UI or relying exclusively on the CLI. This metric, along with data on how many users are using the UI, and whether users are running builds with all required tasks of all required builds vs. builds with some curated selection of tasks, would be interesting to look at to gauge usability of the patch build configuration process. I believe we already can get a sense of how many users are using the UI to configure patch builds by looking at the POST requests to the patch build configuration page. As a follow-up question related to this feature request, are we currently tracking any information on the composition of tasks being selected for patch builds?""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8068","05/26/2020 20:48:51",2,"Child task of display task has ETA based on whole display task's makespan instead of the previous execution of the child task ""When viewing the child task of a display task, such as [this one|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_multiversion_sharding_jscore_multiversion_passthrough_gen_patch_c480627eae1b5d09f8eea247b396b45b89a5c275_5ecd672a306615532c0d3e5b_20_05_26_19_00_10/1##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%2522c480627eae1b5d09f8eea247b396b45b89a5c275%2522%257D%255D%257D], I expect the ETA to be 2-3 minutes based on execution 0. Instead when it's running, the ETA is 16 minutes long, as demonstrated by the attached screenshot. My guess is that child task is using the parent display task's previous makespan (which was 17 minutes) to make this estimation. (Parent of the above linked task [is here)|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_multiversion_display_sharding_jscore_multiversion_passthrough_patch_c480627eae1b5d09f8eea247b396b45b89a5c275_5ecd672a306615532c0d3e5b_20_05_26_19_00_10##%257B%2522compare%2522%253A%255B%257B%2522hash%2522%253A%2522c480627eae1b5d09f8eea247b396b45b89a5c275%2522%257D%255D%257D]""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8071","05/26/2020 22:14:14",1,"The CLI's host create command's key parameter is unclear ""The key help string says """"name or value of an public key to use"""" but it really means the name of a key in the user's user document. The error message for a non-existent key is also not helpful: """"key does not start with ssh-rsa or ssh-dss"""". We should say the key doesn't exist.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9210","05/26/2020 22:32:16",2,"New repobuilder jobs should clear .olddata from previous runs ""When the app server restarts in the middle of a repobuilder job, it will leave behind the {{.olddata}} directory if it was running {{createrepo}} when it restarted. This will cause future {{createrepo}} commands in later repobuilder jobs to fail since the old data directory still exists. We should remove it before we run {{createrepo}} (this should be fine to do since the jobs are protected by scopes on the distro and have to run serially on each distro). Logs: https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Dbarque%20build-repo.distro.rpm.repo.c0860eb1-cce1-4207-8f47-50c786814d87&display.page.search.mode=verbose&dispatch.sample_ratio=1&earliest=1590513780&latest=1590528225&sid=1590528302.2773197""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8075","05/27/2020 15:47:14",2,"Create Query to fetch correct download links for download card ""[https://github.com/evergreen-ci/evergreen/blob/master/environment.go#L736]""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8076","05/27/2020 16:02:41",2,"Create Query to query for correct version of evergreen cli ""https://github.com/evergreen-ci/spruce/pull/179#discussion_r430726473""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8358","05/27/2020 21:06:47",0,"subprocess.exec is missing json tags ""It won't serialize properly if the JSON tags aren't included because it will use the field name starting with an uppercase as the key.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9291","05/28/2020 16:04:27",1,"Allow number of tasks for task group to be configurable ""Although the lower limit is 4, we should let people set it to a different value than that if they want.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8082","05/28/2020 18:39:49",2,"Slack Patch Notification reports failed task has started ""I received a notification (Subscription: 5ec56ff961837d675d29296e; Event: 5ecff63f7742ae725a5508e2) for [this|https://evergreen.mongodb.com/task/jstestfuzz_self_tests_query_agg_fuzzers_v4.4_aggregation_multiversion_fuzzer_inMem_patch_a8b108b061d064bbeebd5bdb7d81ced4d77e1d19_5ecff45a1e2d17733479353c_20_05_28_17_26_51/0] task, see attached screenshot. It failed, the line to the left is red (as it should be in failed scenarios) but the status has been reported as started. Also, this is a PR notification, so I'm not sure why I'm getting a notification for this task, which I never subscribed to, instead of the patch as a whole.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8846","05/28/2020 21:33:45",2,"Reorganize build files ""While the multi-file support is correct and usable, it's probably not the ideal file organization structure. It would be nice to figure out a file organization that somehow nicely handles evergreen-specific configuration (e.g. loggers and task groups) and generator-specific configuration (e.g. Golang's {{RootPackage}} doesn't have a good location, nor do the environment variables) without creating a proliferation of lots of small single-purpose files.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8111","05/29/2020 14:25:25",2,"Not authing against corp secure properly when using performance analysis and triage service ""It appears that the performance analysis and triage service in production is sometimes unavailable. Reproduce by following this [link|https://evergreen.mongodb.com/perf-bb/expanded-metrics-change-points/sys-perf]. Be sure to clear cookies [here|https://signal-processing-service.server-tig.prod.corp.mongodb.com/docs] first. As a DAG engineer, I'd like this to work, and redirect users to authenticate through corpsecure when it doesn't, such that, we have the necessary auth cookies to make these requests. AC: * Evergreen auths against corpsecure as needed""","",0,1,0,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8083","05/29/2020 18:09:39",2,"Setup Failure: test-cloud on Ubuntu 16.04 [Evergreen @ 51e5e1e8] "" h2. [test-cloud failed on Ubuntu 16.04|https://evergreen.mongodb.com/task/evergreen_ubuntu1604_test_cloud_51e5e1e8c347d72879f859868b5d549469ecd6ab_20_05_29_14_35_25/0] Host: [ec2-3-91-191-170.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-09bcb395f9d9879be] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/51e5e1e8c347d72879f859868b5d549469ecd6ab]: Revert """"Revert """"Revert """"Revert """"MAKE-1152: Use MaxInt32 instead of MaxInt64 (#3621)"""""""""""""""" This reverts commit 0c48a16288bb3eeb9cbd80af9c8a5869f9a57d22. | 29 May 20 14:35 UTC Evergreen Subscription: ; Evergreen Event: [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_ubuntu1604_test_cloud_51e5e1e8c347d72879f859868b5d549469ecd6ab_20_05_29_14_35_25/0?type=T] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-8084","05/29/2020 19:00:54",0,"Host should decommission once max provision attempts is exceeded ""There's an off by one error where, because we increment the provision count at the end of the job, the condition {{shouldRetryProvisioning()}} will never return false, so the host will not decommission immediately when it runs out of attempts (it'll be caught by the 25 minute timeout, but it would be better for it to fail fast).""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8985","05/29/2020 21:02:06",8,"Implement jasper build ""The build system should be able to run targets and should be integrated with the amboy job queues/dependency manager to handle parallel target creation/re-building.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8343","05/29/2020 21:03:59",1,"Update signal processing algorithms readme (poetry) ""As a DAG engineer, I'd like to update the signal processing algirithm's readme to reflect the move to pytest, such that, developers know how to develop the library.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8087","05/31/2020 04:09:15",1,"Raise limit on hosts.create ""For provider: docker it seems like it could be much higher.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9163","06/01/2020 16:06:09",2,"Scoped jobs that are undispatchable should have scopes removed or be deleted ""It seems like barque still has problems with re-dispatching old jobs with scopes that exceed the lock timeout. One possible reason is that jobs with a {{DispatchBy}} requirement may not be able to run if too long has passed, so they will be stuck unable to dispatch and also will not clear their scopes. For jobs that have scopes and are stuck in progress but can't be re-dispatched due to {{DispatchBy}}, we should just unset their scopes and mark them complete.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8983","06/01/2020 16:45:58",0,"Move file checksum logic to utility ""This is mostly so that jasper can re-use it for an experimental dependency cache implementation.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8096","06/01/2020 19:07:59",2,"Task spawn host button misses artifacts ""When I try to spawn a host for my BF that had a crash on https://evergreen.mongodb.com/task/mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_large_txns_format_concurrency_sharded_with_stepdowns_202f27642004fc8a87efa543940825b4f5c8d866_20_05_27_15_56_43/0, I don't get the core dumps downloaded and have to download them myself.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8421","06/01/2020 19:18:51",1,"Benchmarks do not compile with change to loggers ""{{go build ./benchmarks}} fails because the loggers were refactored. We should also add it to the lint packages so it can be sanity check that it actually works in the future.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8099","06/01/2020 19:51:11",2,"Improve event log for deleted commit queue item ""When a commit queue item is deleted and the merge is prevented, it may not be immediately clear to the user or to others why the merge task is blacklisted (see [example|https://evergreen.mongodb.com/task/mongodb_mongo_v4.2_commit_queue_merge_merge_patch_patch_495ea4a50219e08ac7a4e1bcd58dc4935f5f0173_5ed53e152fbabe33b5eb87e6_20_06_01_17_43_05#/log/EV#comparehashes=495ea4a50219e08ac7a4e1bcd58dc4935f5f0173&threads=maxonly]). We should add a new type of log that says something like """"Patch deleted by """" before logging that mci has changed the task priority. [https://github.com/evergreen-ci/evergreen/blob/1f42d4e8c3ec4e8b9f1104668ccf41d8040b7f0e/model/commitqueue/commit_queue.go#L271]""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8100","06/01/2020 21:43:57",2,"Can't spawn host after creating a 700GB Volume ""Once I have a workstation with a 700GB volume, the spawn button is greyed out regardless of which distro I try to spin up""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8102","06/02/2020 02:32:55",0,"Fix unit tests ""Because tychoish :( ""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8103","06/02/2020 14:08:13",2,"Spawning containers from tasks shouldn't use existing parent hosts ""The parents will have the same security group as the one running the task, and hosts within this same security group cannot communicate with each other""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8105","06/02/2020 14:52:04",1,"Patch & Task page: set max width on display name part of tables ""when display names are one word and long, there is not enough space to view the other columns""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8106","06/02/2020 14:56:57",1,"GraphQL error: internal system error on logs page ""For https://spruce.mongodb.com/task/evergreen_lint_generate_lint_patch_f6ffa2cce21a18c650c8e61dfca474a1f87845a5_5eb2d84c5623433079333f22_20_05_06_15_31_26/logs. Looking at the legacy UI, this task has no logs, so I think this is just a case where the message should say """"No logs found"""" or something similar instead of the error message.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8107","06/02/2020 14:59:20",2,"Patches that will not run are marked running on patches page in Spruce ""This patch is marked as running on my my patches page https://spruce.mongodb.com/patch/5eb2d84c5623433079333f22/tasks.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8108","06/02/2020 14:59:51",2,"host.create container logs not working ""I wrote a command that tries to tail the logs but they're always empty""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8109","06/02/2020 15:01:01",2,"Malformed link to jira ticket for revert commits on waterfall page ""{color:#1d1c1d}On waterfall page for project {color}{color:#1d1c1d}evergreen is generating wrong URLs for revert commits. In particular is including a double apices {color}{{""""}}{color:#1d1c1d} and colon in the jira [link|https://jira.mongodb.org/browse/%22SERVER-50251:], like the following: !Screen Shot 2020-08-24 at 1.06.36 PM.png|width=30%! ""","",0,0,1,0,0,0,0,0,0,0,1,0,0,0,0
+"EVG-8112","06/02/2020 17:50:11",1,"Task action items in dropdown menu are unresponsive to clicks ""Task page, task actions dropdown menu, clicking on the dropdown items does nothing. Update: the dropdown items do work, but they are not being properly displayed when they are disabled. They are disabled but do not appear disabled; therefore confusing the user by making them think the button is broken. ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8113","06/02/2020 18:02:49",1,"Project setup script doesn't run correctly ""1) The working directory is set to a directory that doesn't exist. 2) The same command runs multiple times.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8833","06/02/2020 19:31:22",2,"Support test results parsing from generators ""The Make-based generator has to have a means to report results from tests (e.g. {{gotest.parse_files}}). The golang generator will also need to support a means to report test results in such a way that we don't need an intermediate file containing the results, but this can occur within subprocess.scripting once test results are moved to cedar.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9178","06/02/2020 20:26:24",1,"Add documentation to exported fields in cedar ""Some of the structs and public fields in Cedar don't have documentation associated with them. It should be added in the places it is needed for the things that the DAG team owns - Performance Analysis Service, time series data, time series update job, etc AC: * Public fields in DAG related areas in Cedar are documented""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8119","06/02/2020 21:18:16",1,"refactor and Remove XHR dependency for commit_queue tests ""We use a pollyfill with in our cypress tests to listen to our graphql requests. This breaks apollos ability to make event driven callbacks like `onError` and causes functionality to break within cypress but not on a normal browser. This ticket covers removing all `waitForGQL` dependencies with in the commit_queue tests""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8133","06/02/2020 21:56:52",1,"""Unable to find task"" when clicking on burn_in task ""Clicking the task shown in the screen shot for this [patch build|https://spruce.mongodb.com/patch/5ea9c9fc7742ae2842c23bb3/tasks] leads to the graphql error in the screen shot. Here is the [offending task|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_burn_in:replica_sets_jscore_passthrough_1_enterprise_rhel_62_64_bit_patch_4fa07f1c28e20f902d55dcc2f2c07a8135ad31f4_5ea9c9fc7742ae2842c23bb3_20_04_29_18_40_22/0##comparehashes=4fa07f1c28e20f902d55dcc2f2c07a8135ad31f4&threads=all] in the legacy UI.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8134","06/02/2020 22:47:16",2,"Expose --add-host for Docker containers created with host.create ""It would be useful to be able to say something like this: {code:yaml} - command: shell.exec params: script: | echo """"task_host_server_ip: $(ip addr show ...)"""" > task_host_expansions.yml # or the external IP from the instance metadata service - command: expansions.update params: file: task_host_expansions.yml - command: host.create params: provider: docker image: selenium/standalone-chrome publish_ports: true add_hosts: task_host: """"${task_host_server_ip}"""" {code} Then within the container, the browser could be directed to visit http://task_host:9080. Reference: https://docs.docker.com/engine/reference/run/#managing-etchosts"""," - command: shell.exec params: script: | echo """"task_host_server_ip: $(ip addr show ...)"""" > task_host_expansions.yml # or the external IP from the instance metadata service - command: expansions.update params: file: task_host_expansions.yml - command: host.create params: provider: docker image: selenium/standalone-chrome publish_ports: true add_hosts: task_host: """"${task_host_server_ip}"""" ",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8135","06/03/2020 14:20:34",1,"my patches: fix build status icon link ""currently uses the wrong id""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8136","06/03/2020 15:13:20",2,"Add count of how long task has been running / ETA ""Currently there's a count up of how long the task has been running, and and count down for ETA. This is prominently featured in the old UI and would be missed in the new UI. Also on tasks that haven't started yet we show the estimated time to start and the estimated duration. !image-2020-06-03-07-13-11-390.png!""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8138","06/03/2020 15:16:25",1,"Event logs aren't handling quotes correctly ""[https://spruce.mongodb.com/task/evergreen_ubuntu1604_dist_patch_40de312f8fb0855a0d4fcaf1647b36fc6173768b_5ed6d4df850e612690f882f0_20_06_02_22_38_41/logs?logtype=event]""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8140","06/03/2020 15:26:51",2,"Distinguish between Undispatched and Blocked tasks in Spruce ""This task shows undispatched, but it is blocked on other tasks succeeding which have failed so the status should say blocked. [https://spruce.mongodb.com/task/evergreen_lint_lint_trigger_patch_1f42d4e8c3ec4e8b9f1104668ccf41d8040b7f0e_5ed671e1d6d80a60ea7b2677_20_06_02_15_36_32/logs]""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8142","06/03/2020 15:33:19",2,"Task durations seem wrong ""The point of showing the duration on the current task page is to show the length of time that the current task takes, not the overall time of the commit. In [Spruce|https://spruce.mongodb.com/task/evergreen_lint_lint_rest_route_patch_1f42d4e8c3ec4e8b9f1104668ccf41d8040b7f0e_5ed671e1d6d80a60ea7b2677_20_06_02_15_36_32/logs] this task duration is 13 hours, and the base commit duration is 7 hours. But [in Evergreen |https://evergreen.mongodb.com/task/evergreen_lint_lint_rest_route_patch_1f42d4e8c3ec4e8b9f1104668ccf41d8040b7f0e_5ed671e1d6d80a60ea7b2677_20_06_02_15_36_32]the task duration is 47 seconds, with 27 seconds on the base commit, which is much more relevant to the current task. ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8143","06/03/2020 15:33:42",2,"Patch shows failed, but no tasks are failed ""https://spruce.mongodb.com/patch/5ed7b0670ae606496cf87284/tasks?variant=ubuntu1604 Note that this appeared to be the case before test-cloud ran. The icon for the build variant on the my patches page was correct (that the variant was in progress) but the patch badge said failed.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8144","06/03/2020 15:35:06",2,"Is anything actually ever in Running status? ""Tasks that are running appear to be Started. I don't know if anything is ever in Running. Is that status redundant?""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8145","06/03/2020 15:37:20",1,"Time taken missing, though makespan is present ""https://spruce.mongodb.com/patch/5ed7b0670ae606496cf87284/tasks?variant=ubuntu1604""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8146","06/03/2020 15:40:42",1,"Variant should use display name, not id ""For example, https://spruce.mongodb.com/patch/5ed7b0670ae606496cf87284/tasks?page=0 should have """"Ubuntu 16.04"""", not """"ubuntu1604"""" in both the table and the build variants pane. That should also be the case for the tooltips on the my patches page.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8151","06/03/2020 18:53:20",2,"subprocess.scripting should parse go test results ""We should not have to execute a separate {{gotest.parse_files}} command when it could be done directly from the output of subprocess.scripting. This is just an intermediate solution until we move test results to cedar.""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8547","06/03/2020 22:55:31",0,"gotest.parse commands need to specify files ""We can't specify the files to parse for golang tests currently.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8157","06/04/2020 14:21:44",1,"Two Linux 64 bits show up on the download card ""!image-2020-06-04-09-20-57-145.png! One of the them is the legacy version we shouldn't show that""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8360","06/04/2020 16:48:50",0,"Return correct gotest command name based on legacy or non-legacy ""It returns the wrong command name if you use {{gotest.parse_files}} instead of {{gotest.parse_json}}""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9282","06/04/2020 17:18:57",1,"Experiment with removing grep for go test ""I feel like extra grepping for pass/fail in the output log is unnecessary if you just check the return code of {{go test}}.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9005","06/04/2020 17:41:02",1,"Curling golangci-lint should retry ""Sometimes the host fails to resolve {{raw.githubusercontent.com}}.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8159","06/04/2020 17:54:24",1,"gotest.parse_json should be able to match file glob patterns ""File path matching behavior should be the same as {{gotest.parse_files}} basically.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8507","06/04/2020 18:57:36",1,"Fix test-remote on race-detector ""The test only fails on race detector so it could be something to do with the test execution being slower due to race detection enabled. Logs: https://evergreen.mongodb.com/task/jasper_race_detector_test_remote_edc31da19da5dea040847c2c83e034fb2a357cd7_20_06_02_14_25_50""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8163","06/05/2020 13:59:36",2,"Patch base task statuses filters do not always work ""On the patch details page, filtering by base status returns unexpected results. Sometimes, the statuses displayed in the filter dropdown do not align with the statuses displayed in the base status column. Additionally, filtering by 'success' or another status returns no tasks when it should return some or all of them. Example patch: [https://spruce.mongodb.com/version/5ecff20932f4173f1fdc1666/tasks?limit=100&page=0]""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8164","06/05/2020 14:24:21",2,"Why did git.get_project block the mongodb-mongo-master commit queue for 35 minutes? ""This morning I created a {{commit-queue merge}} ([here|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_lint_fuzzer_sanity_patch_patch_5c6e1e540a9b5b0e0c87fa3522b9e668e2357a8c_5eda3142d6d80a270f66d4cf_20_06_05_11_50_01##comparehashes=5c6e1e540a9b5b0e0c87fa3522b9e668e2357a8c&threads=maxonly]) which locked up for 35 minutes. (compared to the [expected 2 minutes|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_lint_fuzzer_sanity_patch_patch_7074dee1fbf6763c0d463c377c2e47d8ef2c4f6f_5eda44a72a60ed0ac4af4b34_20_06_05_13_12_08##comparehashes=7074dee1fbf6763c0d463c377c2e47d8ef2c4f6f&threads=maxonly] for this task) There is no output from {{git.get_project}}. I suspect this would've blocked the commit queue until it timed out. Can you look into whether this was an Evergreen, VPC/networking issue, Github or other issue? Furthermore, what can be done to prevent this in the future? Should we set timeouts for the commit queue tasks in the server yaml. Does the git command support a timeout for the initial round trip connection (which I'm assuming is what failed here)? Basically, I could've tanked the whole queue for however long the timeout is, and that would've been pretty bad for others.""","",0,0,0,0,0,1,0,0,0,1,0,0,0,0,0
+"EVG-8166","06/05/2020 14:57:30",0,"Remove condition to stop polling build variants when all tasks have finished ""this condition prevents polling when we actually should be""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8167","06/05/2020 15:17:42",2,"Allow passing arbitrary additional parameters to host rsync command ""So that people can do things like pass a .gitignore to apply the rules for file exclusion via {{--filter}} or do whatever they want if they already know how to use rsync.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8170","06/05/2020 17:15:36",0,"Remove notify me button on patch page ""temporarily remove notify me button on patch page until the modal is developed""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8171","06/05/2020 17:33:35",1,"Base commit link on patch details page should go to base commit in legacy ui ""Currently it just goes to the version page""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-9058","06/05/2020 18:52:33",1,"Export scripting harness type strings ""The only way to know the valid harness type strings is to hard-code strings. This should also be fixed in {{subprocess.scripting}}.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8177","06/05/2020 19:38:51",1,"Navigating to certain Spruce pages brings users to the bottom of the screen ""This happens with the patch [configure|https://spruce.mongodb.com/patch/5eda79550ae6061360113a0f/configure/tasks] page: !patch configure page loads starting at the bottom by default.png|thumbnail! as well as the patch details page, when accessed via the breadcrumb. E.g., clicking on {{aggregation_auth}} from [this patch|https://spruce.mongodb.com/version/5ea9c9fc7742ae2842c23bb3/tasks] and then clicking """"Patch 940"""": !brought to the bottom of the page on page load.png|thumbnail! It would be great if these pages went to the top of the page by default.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8178","06/05/2020 19:41:40",1,"Navigating to ""RAW"" link in patch configuration does not open new window and clears task selections ""When I am in the patch configuration window and I view the RAW link for the changes in the patch, my task selection is cleared, since the RAW link does not open the page in a new tab. !clicking on raw logs loses my selected tasks.png|thumbnail! I would expect the RAW link to open the page in a new tab, as is what happens in the HTML page.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8182","06/05/2020 21:00:33",2,"Don't Log IDE panics ""[reverse_proxy|https://github.com/golang/go/blob/cc8838d645b2b7026c1f3aaceb011775c5ca3a08/src/net/http/httputil/reverseproxy.go#L299] sometimes panics if there's an error while streaming the response back to the client (https://github.com/golang/go/issues/23643). Our [middleware|https://github.com/evergreen-ci/gimlet/blob/185166ff85e8801545747fa65a14486e0fe4680b/middleware_grip.go#L166] recovers the panic and logs to slack. Because the panic isn't actionable we need to log at a lower level. One design is to recover the panic in the route so it doesn't make it all the way back up to the recovery logger.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8183","06/05/2020 21:22:51",2,"Give MANA an API to cleanup unexpirable hosts/volumes in off-boarding ""Expose an API for MANA to call in their off-boarding process that removes the user's unexpirable hosts and volumes""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8977","06/05/2020 23:09:05",2,"Allow environment variables to be specified at package and variant levels for golang generator ""For example, we may want to use a different GOROOT for some variants.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8188","06/06/2020 04:17:02",2,"""T-shaped"" patch builds should deduplicate tasks and select tasks on all relevant build variants ""It looks like there are multiple checkboxes for the same task and while both become checked when one is clicked, the task count for only one of the build variants is updated (depending on which checkbox is clicked). Note that I didn't attempt to schedule this patch build to see if the task would actually run on all the build variants or not. Suppose the user takes the following actions on the configure patch build page: # Selects the build variant BV1. # Checks the checkbox for the task T1. # Selects the build variant BV2 which also supports running task T1. The expected state of the checkbox for the task T1 is for it be [partially selected|https://developer.mozilla.org/en-US/docs/Web/HTML/Element/input/checkbox#Indeterminate_state_checkboxes]. Clicking the checkbox while it is in an indeterminate state deselects it.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8189","06/06/2020 04:17:20",2,"""T-shaped"" patch builds should support shift+click ""EVG-7974 added support for cmd+click. shift+click is additionally supported by the existing patch configuration page to select a range of contiguous build variants.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8190","06/06/2020 04:18:01",2,"""T-shaped"" patch builds should mark checkboxes as partially selected (indeterminate) ""Suppose the user takes the following actions on the configure patch build page: # Selects the build variant BV1. # Checks the checkbox for the task T1. # Selects the build variant BV2 which also supports running task T1. The expected state of the checkbox for the task T1 is for it be [partially selected|https://developer.mozilla.org/en-US/docs/Web/HTML/Element/input/checkbox#Indeterminate_state_checkboxes]. Clicking the checkbox while it is in an indeterminate state deselects it.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8930","06/08/2020 13:40:49",2,"BuildFileList should support file exclusion ""We have a pattern of building file lists within a directory if they match a series of include/exclude patterns. We should be able to generalize {{BuildFileList}} so that it can support file inclusion/exclusion rules within a directory (or maybe without specifying a directory).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8960","06/08/2020 13:44:31",2,"Golang package discovery should be able to find source files ""Currently, it will only discover directories containing test files. For one example use case, a user who wants to lint might want a package to be auto-defined if it contains go source files, even if it does not contain any test files. This should be a configurable setting (by default, off).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8192","06/08/2020 14:09:46",0,"Variables should be optional in strict project yaml validation ""People seem to conventionally use it for defining extra yaml stuff like anchors, but it's not required as part of the project config.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9003","06/08/2020 15:14:20",0,"Call git.get_project in setup group ""git.get_project should be done at the beginning of each task group when the host is setting it up but doesn't need to be done for every single task.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9201","06/08/2020 15:15:13",0,"Allow users to specify variables section ""This is so that people can define things like anchors.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8197","06/08/2020 15:34:40",2,"GDB improvements ""Add ~/.gdbinit to * source ~/mongo/.gdbinit and ~/mongo-v44/.gdbinit when invoked from those directories * set these settings: ** Add /opt/mongodbtoolchain/gdb/bin to PATH"""," add-auto-load-safe-path / set pagination off set print object on set scheduler-locking on #set print static-members off #set print elements 0 ",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8198","06/08/2020 15:39:48",2,"Documentation Improvements ""bq. Could be good to have a warning against using ninja for compiledb because I tried that out of habit but it actually doesn’t work right if you built ninja with icecream support bq. I’d appreciate some public documentation on icecream scheduling in the cloud. It’s disconcerting to have it silently populated in /etc/icecc/icecc.conf from the variant image bq. Highlight how Evergreen IDE doesn’t require being on VPN. Go one step further and also suggest using the integrated terminal in VS Code to do server builds, patch builds, commit queue, etc. without being on the VPN.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8200","06/08/2020 17:27:24",1,"No patches listed on spruce patches page ""The following page: https://spruce.mongodb.com/user/mark.benvenuto/patches, displays no patches for me. I expect it would display some patches. !image-2020-06-08-12-26-01-860.png|thumbnail! Browser: Version 81.0.4044.138 (Official Build) (64-bit) OS: macOS 10.14.6 ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8204","06/08/2020 19:40:39",1,"Make virtual workstation setup include code review tool posting to JIRA ""The alias virtual workstations provide for uploading code reviews is cr='~/kernel-tools/codereview/upload.py --check-clang-format --check-eslint'. Our server onboarding instructions include the --jira_user option, so that running the code review tool posts to JIRA. This also requires setting up Jira OAuth credentials. It would be helpful to include this in either the workstation configuration or in the wiki.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8206","06/08/2020 20:54:53",2,"Be more careful about retrying on github 404s ""If a user adds a module that with a typo in the GitHub, Evergreen will exhaust its GitHub API limit retrying. https://evergreen.mongodb.com/version/5ede8b903e8e865991924ea7""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8209","06/09/2020 16:36:40",2,"explain problem building container image errors ""Understand the root cause of """"problem building container image"""" errors""","",0,0,0,0,0,1,0,0,0,1,0,0,0,0,0
+"EVG-8211","06/09/2020 19:45:59",2,"Write Design Document for Parameterized Builds ""Scope: [https://docs.google.com/document/d/1GuZOGa07dMhC1KS8TQeF3oad0Tg5Xp1XleodrcdB2nM/edit#]""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8214","06/10/2020 15:40:47",1,"evergreen local user settings is wrong ""Currently the user settings say: However this doesn't work unless it's """"http://localhost"""" so this should be fixed."""," api_server_host: """"localhost:9090/api"""" ui_server_host: """"localhost:9090""""",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8216","06/10/2020 17:17:28",1,"Add one repo/branch restraint to git tags ""PR testing and commit queues seem restricted to one repo/branch. Git tag triggered versions should likely also be restricted.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9326","06/10/2020 18:52:02",2,"Implement by TaskID route ""Implement REST route for fetching test results by task id and execution.""","",0,0,0,0,0,0,0,0,1,0,0,0,1,0,0
+"EVG-9327","06/10/2020 18:53:00",2,"Implement by TestName route ""Implement REST route to fetch test results by task id, execution, and test name.""","",0,0,0,0,0,0,0,0,1,0,0,0,1,0,0
+"EVG-9316","06/10/2020 19:34:02",1,"Define data model for stats collection ""We need to define the structure of the payloads we are going to write to s3 and that the agent will send over gRPC, as was done [here|https://github.com/evergreen-ci/cedar/blob/master/model/test_results.go#L325-L338].""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8218","06/10/2020 19:43:52",2,"Spruce deploy script should block if local master is not latest ""the script should be modified to prevent pushing to production if the local master branch is not the most recent""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8223","06/10/2020 20:41:10",0,"Implement disk usage stats collector ""This should be a replacement for {{df -h}} in the system logs currently.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8224","06/10/2020 20:47:07",2,"Unhandled error is thrown when spruce disconnects from evergreen app server ""spruce should handle this error so the app doesn't crash, maybe indicate to the user that they have lost connectivity ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8226","06/10/2020 22:12:46",1,"Swap order of base and patch status columns ""Should reflect the legacy ui""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8230","06/11/2020 15:22:46",2,"Users who haven't signed into the UI recently get 401 errors for the CLI ""I assumed that background reauth would have handled this, so I opened this as a bug.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8236","06/11/2020 18:14:31",2,"'My patches' breadcrumb on another user's patch leads to my own patches ""I pasted the version ID of another user's evergreen patch into a Spruce version URL: https://spruce.mongodb.com/version/5ee2434732f417796e65772f/tasks Navigating to 'My Patches' from this other user's patch leads me to my own patches, instead of the user's patches. When I try to hit 'back' to get back to the user's patch link, I remain on my own 'My Patches' page.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8238","06/11/2020 19:35:43",1,"Patch build notification email links to legacy UI patch instead of Spruce UI ""Example: Hi, Your Evergreen patch in 'mongodb-mongo-master' [5ee254c37742ae4114585d57|https://evergreen.mongodb.com/version/5ee254c37742ae4114585d57] has failed. test patch""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8239","06/11/2020 20:00:31",1,"Link to legacy top nav pages in Spruce UI ""In order to allow folks to maintain their navigation patterns to pages not supported in Spruce, we should allow the Spruce top nav to link to the legacy pages: !Screen Shot 2020-06-11 at 2.57.31 PM.png|thumbnail! ""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8242","06/11/2020 22:43:41",2,"Task logs not being rendered properly ""Specially formatted task logs dont have appropriate styling associated with them [https://spruce.mongodb.com/task/spruce_ubuntu1604_e2e_test_patch_ac763049d8990383f90704ded0aa7c1de30cb6a5_5ee2a096a4cf477cceaa2f69_20_06_11_21_22_31/logs?logtype=task] [https://evergreen.mongodb.com/task_log_raw/spruce_ubuntu1604_e2e_test_patch_ac763049d8990383f90704ded0aa7c1de30cb6a5_5ee2a096a4cf477cceaa2f69_20_06_11_21_22_31/0?type=T] !https://files.slack.com/files-pri/T024FNNHU-F015N84P51P/image.png! !https://files.slack.com/files-pri/T024FNNHU-F0159RC0S59/image.png!""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8250","06/15/2020 11:15:30",2,"Unkown directive client bugsnag error ""Bugsnag is throwing graphql validation errors when using the @client directive in graphql queries. Either replace the usage of this field or handle this error. *[https://app.bugsnag.com/mongodb/evergreen/errors/5ede8fc73dc90800185da435?filters[event.since][0]=30d&filters[error.status][0]=open&pivot_tab=event]*""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8252","06/15/2020 13:57:51",1,"Database error retrieving DistroQueueInfo for distro id 'suse15-zseries-large': document not found ""I see these errors in Splunk. This is likely a side effect of the work to remove queue distros when scheduling is disabled.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-8253","06/15/2020 14:22:17",1,"Version start time should not get updated with every task ""https://github.com/evergreen-ci/evergreen/blob/bd4afd4af6d8ea1e5b3bf7abfc8955170969db1d/service/api_task.go#L53 will eventually call into a function that updates the version start time regardless of if it's already started""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8255","06/15/2020 15:20:20",2,"Login cache should be cleared as part of mana offboarding ""MANA is expected to call `DELETE /users/{user_id}/permissions` in order to clear the user's permissions, but it would be good to also clear their login cache so that the background reauthentication job does not needlessly try to reauth people who should already be offboarded. We should either add add another route and have MANA make a request to it, or just add it to the existing delete permissions route.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8259","06/15/2020 16:20:55",0,"Jasper restart should be allowed even if host is in running state ""Some of the mac hosts were restarted but the agent was stuck so they never got put in the 'provisioning' state. Usually, we want them to be in the provisioning state rather than running state because it guarantees that the agent monitor has shut down. However, [this check for the agent monitor|https://github.com/evergreen-ci/evergreen/blob/bd4afd4af6d8ea1e5b3bf7abfc8955170969db1d/units/provisioning_jasper_restart.go#L139-L149] should be sufficient to prevent jasper restarts until the agent monitor is down. Therefore, it should be fine to allow hosts to be in the running state when jasper needs to be restarted. Logs: https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Devergreen%20macos-1014-57.macstadium.build.10gen.cc%20NOT%20external-state-check%20NOT%20%2Fhost%2Fmacos-1014-57.macstadium.build.10gen.cc&display.page.search.mode=verbose&dispatch.sample_ratio=1&earliest=1591887600&latest=1591920000&sid=1592233844.3764706""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8260","06/15/2020 16:48:08",1,"display spruce git commit hash in production html ""this will help us keep track of what is deployed to production""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8261","06/15/2020 17:54:04",2,"Store page size in local storage and use as default in Spruce ""use the latest page size used (e.g., 20 [entries per page|https://spruce.mongodb.com/version/5efa5aac2fbabe4f8783c596/tasks?limit=20&page=0], 100 [entries per page|https://spruce.mongodb.com/version/5efa5aac2fbabe4f8783c596/tasks?limit=100&page=0], etc.) as the default value when limit query variable does not exist in url ""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8270","06/15/2020 22:44:22",1,"adjust welcome modal styles ""add space between description and carousel navigation dots and center the modal""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8282","06/16/2020 18:46:05",2,"Display tasks have duplicate dependencies ""EVG-6917 was supposed to make display tasks not have duplicate dependencies, but they still appear to. e.g. https://evergreen.mongodb.com/task/mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_display_rollback_fuzzer_unclean_shutdowns_patch_40add1dc44736475ea3aa598f9878e990b2ad974_5ee890a21e2d17016ea3405d_20_06_16_09_29_04""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8284","06/16/2020 19:49:18",1,"Remove multiple AWS key/secrets from admin settings ""Turns out this is unnecessary, all regions will use the same key/secret pair.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8293","06/17/2020 04:05:44",2,"Improve display of long variant/test names in Spruce ""see https://spruce.mongodb.com/version/5ee93748562343577c5e7870/tasks?limit=100&page=0 and https://spruce.mongodb.com/task/mongo_java_driver_tests_jdk_secure__version~latest_os~linux_topology~sharded_cluster_auth~auth_ssl~ssl_jdk~jdk11_test_69ef5be50c8ac47a989dcf92f230c837f4454df3_20_06_15_16_44_05/tests?page=0&testname=ServerDiscovery""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8299","06/17/2020 16:18:44",2,"CLI command to ssh to an instance ""Wrap ssh to look up the current DNS name of the instance so a user can always use to same command""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8300","06/17/2020 17:05:41",2,"generate.task timing out ""It seems like there have been a lot of 'generate.tasks' calls that are timing out at 2 hours. They feels like a really long time for 'generate.tasks' to run. It might be worth investigating if something is wrong.""","",0,0,0,0,0,0,1,0,0,1,0,0,0,0,0
+"EVG-8310","06/18/2020 14:25:50",2,"Makespan is skewed for patches that have had tasks restarted ""I re-ran one of the tasks in one of my [patches|https://spruce.mongodb.com/version/5ee254c37742ae4114585d57/tasks] a few days after initially submitting it. The Makespan includes the multiple days worth of time between the initial patch start time, and the finish time of the second execution of the tasks: !Screen Shot 2020-06-18 at 9.18.50 AM.png|width=30%! This behavior is consistent with that of the [legacy UI|https://evergreen.mongodb.com/version/5ee254c37742ae4114585d57]. I would expect that the idle time during which no tasks are running would not be included in the makespan. It would be helpful to consider how to account for this idle time when displaying Makespan / time taken.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-8318","06/18/2020 20:13:46",1,"Link generated tasks to its generator ""There is not currently a way in the UI to map a generated task to its generator. We should expose this in both the legacy and new UIs, since EVG-8191 has a dependency on this ticket.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8320","06/18/2020 20:50:57",2,"Evergreen task ran twice on the same machine (under the same task execution) ""h6. Task logs https://evergreen.mongodb.com/task_log_raw/mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_sharding_auth_audit_13_enterprise_rhel_62_64_bit_d36d5430d7aefc844a2020064703df4fff7d6dcd_20_06_04_12_54_42/0?type=T#L871 h6. Agent logs https://evergreen.mongodb.com/task_log_raw/mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_sharding_auth_audit_13_enterprise_rhel_62_64_bit_d36d5430d7aefc844a2020064703df4fff7d6dcd_20_06_04_12_54_42/0?type=E#L1172 """," [2020/06/04 14:37:47.739] Starting task mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_sharding_auth_audit_13_enterprise_rhel_62_64_bit_d36d5430d7aefc844a2020064703df4fff7d6dcd_20_06_04_12_54_42, execution 0. ... [2020/06/04 15:43:33.183] Starting task mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_sharding_auth_audit_13_enterprise_rhel_62_64_bit_d36d5430d7aefc844a2020064703df4fff7d6dcd_20_06_04_12_54_42, execution 0. [2020/06/04 14:37:47.739] Making new folder for task execution: /data/mci/18a4d6b3d9435b2342c2680aeb8644b4 [2020/06/04 14:37:47.768] Reporting task started. [2020/06/04 14:37:47.869] cleaning up processes for task: mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_sharding_auth_audit_13_enterprise_rhel_62_64_bit_d36d5430d7aefc844a2020064703df4fff7d6dcd_20_06_04_12_54_42 ... [2020/06/04 15:43:33.183] Making new folder for task execution: /data/mci/89747a5447c929f4943629c1d05667ea [2020/06/04 15:43:33.184] Reporting task started. [2020/06/04 15:43:33.326] cleaning up processes for task: mongodb_mongo_v4.4_enterprise_rhel_62_64_bit_sharding_auth_audit_13_enterprise_rhel_62_64_bit_d36d5430d7aefc844a2020064703df4fff7d6dcd_20_06_04_12_54_42 ",0,0,0,0,0,0,1,0,0,1,0,0,0,0,0
+"EVG-8322","06/22/2020 15:09:14",1,"user/patches & user/:id path redirect should replace browser history instead of push ""this is to avoid polluting browser history""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8323","06/22/2020 15:19:37",2,"remove scroll and show all items at once from changes and files tables ""this makes sense since the build variant side panel is not scrollable and is as long as it needs to be""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-8325","06/22/2020 15:32:49",2,"Failures: test-rest-model on Race Detector [Evergreen @ 769da5ca] (TestCodegen, basic, scalars, thirdparty, TestCreateConversionMethodsError) "" h2. [test-rest-model failed on Race Detector|https://evergreen.mongodb.com/task/evergreen_race_detector_test_rest_model_769da5ca31dd3ee3d26720269ecb4d2ceab0808e_20_06_18_20_45_59/0] Host: [ec2-54-226-139-238.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-02ca1699534bb8bf4] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/769da5ca31dd3ee3d26720269ecb4d2ceab0808e]: MAKE-1363: retry curling golangci-lint (#3709) | 18 Jun 20 20:45 UTC Evergreen Subscription: ; Evergreen Event: *TestCodegen* - [Logs|https://evergreen.mongodb.com/test_log/5eebd888a4cf47327ca2e9ed] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_race_detector_test_rest_model_769da5ca31dd3ee3d26720269ecb4d2ceab0808e_20_06_18_20_45_59#TestCodegen=fail] *basic* - [Logs|https://evergreen.mongodb.com/test_log/5eebd888a4cf47327ca2e9ed] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_race_detector_test_rest_model_769da5ca31dd3ee3d26720269ecb4d2ceab0808e_20_06_18_20_45_59#basic=fail] *scalars* - [Logs|https://evergreen.mongodb.com/test_log/5eebd888a4cf47327ca2e9ed] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_race_detector_test_rest_model_769da5ca31dd3ee3d26720269ecb4d2ceab0808e_20_06_18_20_45_59#scalars=fail] *thirdparty* - [Logs|https://evergreen.mongodb.com/test_log/5eebd888a4cf47327ca2e9ed] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_race_detector_test_rest_model_769da5ca31dd3ee3d26720269ecb4d2ceab0808e_20_06_18_20_45_59#thirdparty=fail] *TestCreateConversionMethodsError* - [Logs|https://evergreen.mongodb.com/test_log/5eebd888a4cf47327ca2e9ed] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_race_detector_test_rest_model_769da5ca31dd3ee3d26720269ecb4d2ceab0808e_20_06_18_20_45_59#TestCreateConversionMethodsError=fail] [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_race_detector_test_rest_model_769da5ca31dd3ee3d26720269ecb4d2ceab0808e_20_06_18_20_45_59/0?type=T] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-9689","06/22/2020 18:16:19",2,"Remove timeline and summary pages ""Research has found that there are no users of the timeline and summary pages. We should remove them.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-9690","06/22/2020 20:31:59",1,"Auto-populate working directory if not specified in subprocess.scripting ""If the working directory is not specified for subprocess.scripting, the code most likely will not work. However, most of the time, it seems like using the agent's working directory should be sufficient as an auto-populated field. This will help the jasper generator not need to depend on the host's working directory (which enables local jasper build generation to be passed directly with a patch).""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9700","06/23/2020 17:58:48",2,"Add more robust buildlogger testing for resmoke ""These tests include: * End-to-end test with curator binary. * Test to retrieve grouped logs with disjoint times. * Test to retrieve grouped logs with overlapping tags.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9701","06/23/2020 18:15:53",1,"Passing a config file that doesn't exist should alert the user ""If someone tries to pass a config file that doesn't exits, (evergreen -c ~/.non_existent_file.yml) the error they get is """"can't get project ref: Unexpected reply from server (404 Not Found): no project found named """" It should instead display a more helpful error like “hey that configuration file doesn’t exist”""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9703","06/23/2020 20:04:55",2,"Commit queue not working for sandbox in staging ""For whatever reason, the commit queue for yb_mci in staging seems to not be working (I'm not sure if it's the commit queue for all of staging or for this project in either environment or what). For example: [this version|https://evergreen-staging.corp.mongodb.com/version/5ef23ae597b1d31c63ef1665] has no merge task associated with it. Having looked in the DB, the parser project associated with the version _does_ have the merge task and variant, but the tasks/builds don't exist. This project in staging is important for our own development of the commit queue so we should resolve this.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9706","06/24/2020 01:30:47",2,"""Out of Memory Kill detected"" message on task page is incomplete for some OOM failures ""(/) """"Out of Memory Kill detected (PIDs: 2762)"""" - https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_ubuntu_task_executor_pool_size_1604_64_bit_concurrency_sharded_replication_57faf552c3acb1dbfacff0020003cbcd3bd36d17_20_06_17_01_34_24/0#/log/S (x) """"Out of Memory Kill detected (PIDs: )"""" - https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_inmem_causally_consistent_read_concern_snapshot_passthrough_0_enterprise_rhel_62_64_bit_inmem_2cc507c7c4199aeddd698a277a346b2acfa1c224_20_06_22_01_14_04/0#/log/S Both of these failures occurred after the changes from [af7fb3b|https://github.com/mongodb/mongo/commit/af7fb3b79ce6ad97b8288c5e7cd64c405375e79b] as part of SERVER-48455."""," [2020/06/17 04:27:33.073] OOM (Out of memory) killed processes detected [2020/06/17 04:27:33.073] [Wed Jun 17 04:05:50 2020] evergreen invoked oom-killer: gfp_mask=0x24201ca, order=0, oom_score_adj=0 [2020/06/17 04:27:33.073] [Wed Jun 17 04:05:50 2020] Out of memory: Kill process 2762 (mongod) score 163 or sacrifice child [2020/06/17 04:27:33.073] [Wed Jun 17 04:05:50 2020] Killed process 2762 (mongod) total-vm:3492232kB, anon-rss:1247060kB, file-rss:0kB [2020/06/22 11:01:26.752] OOM (Out of memory) killed processes detected [2020/06/22 11:01:26.752] js invoked oom-killer: gfp_mask=0x280da, order=0, oom_adj=0, oom_score_adj=0 [2020/06/22 11:01:26.752] Out of memory: Kill process 9784 (mongod) score 121 or sacrifice child [2020/06/22 11:01:26.752] Killed process 9784, UID 0, (mongod) total-vm:3098244kB, anon-rss:1157280kB, file-rss:32kB [2020/06/22 11:01:26.752] ReplCoo.Network invoked oom-killer: gfp_mask=0x201da, order=0, oom_adj=0, oom_score_adj=0 [2020/06/22 11:01:26.752] Out of memory: Kill process 9823 (FlowCon.fresher) score 121 or sacrifice child [2020/06/22 11:01:26.752] Killed process 9823, UID 0, (FlowCon.fresher) total-vm:3098244kB, anon-rss:1157280kB, file-rss:36kB ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-9707","06/24/2020 14:47:15",1,"evergreen YAML should not write AWS credentials to home directory ""It should follow what pail does and export environment variables instead (and remove {{~/.aws}} since we can't guarantee that other projects will clean up after themselves).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-9708","06/24/2020 14:56:44",1,"Failures: test-operations on Windows [Evergreen @ c2718716] (TestHostRsync, ExistingLocalFileMirrorsRemoteFile, NoChangesInDryRunWithPull, RemoteDirectoryHasLocalSubdirectoryWithoutTrailingSlash +9 more) "" h2. [test-operations failed on Windows|https://evergreen.mongodb.com/task/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24/0] Host: [ec2-18-207-164-212.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-0779e1b12bb66e623] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/c2718716d542f59601fc6b5e86cb4b8fcda0b3f1]: EVG-8204 increment version string | 23 Jun 20 21:29 UTC Evergreen Subscription: ; Evergreen Event: *TestHostRsync* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#TestHostRsync=fail] *ExistingLocalFileMirrorsRemoteFile* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#ExistingLocalFileMirrorsRemoteFile=fail] *NoChangesInDryRunWithPull* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#NoChangesInDryRunWithPull=fail] *RemoteDirectoryHasLocalSubdirectoryWithoutTrailingSlash* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#RemoteDirectoryHasLocalSubdirectoryWithoutTrailingSlash=fail] *RemoteDirectoryMirrorsLocalDirectoryWithoutDeletions* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#RemoteDirectoryMirrorsLocalDirectoryWithoutDeletions=fail] *LocalDirectoryMirrorsRemoteDirectoryWithoutDeletions* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#LocalDirectoryMirrorsRemoteDirectoryWithoutDeletions=fail] *RemoteFileCreatedIfNonexistent* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#RemoteFileCreatedIfNonexistent=fail] *ExistingRemoteFileMirrorsLocalFile* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#ExistingRemoteFileMirrorsLocalFile=fail] *NoChangesInDryRun* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#NoChangesInDryRun=fail] *LocalDirectoryMirrorsRemoteDirectoryExactly* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#LocalDirectoryMirrorsRemoteDirectoryExactly=fail] *LocalDirectoryHasRemoteSubdirectoryWithoutTrailingSlash* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#LocalDirectoryHasRemoteSubdirectoryWithoutTrailingSlash=fail] *LocalFileCreatedIfNonexistent* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#LocalFileCreatedIfNonexistent=fail] *RemoteDirectoryMirrorsLocalDirectoryExactly* - [Logs|https://evergreen.mongodb.com/test_log/5ef283540305b93e1b26de4b] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24#RemoteDirectoryMirrorsLocalDirectoryExactly=fail] [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_windows_test_operations_c2718716d542f59601fc6b5e86cb4b8fcda0b3f1_20_06_23_21_29_24/0?type=T] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-9709","06/24/2020 17:10:52",0,"Fix compilation of benchmarks ""Cedar's benchmarks don't compile because the timber client's function signature changed.""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12386","06/25/2020 03:55:25",2,"Add region credentials to distros in staging ""Add support for the following regions in all ( ?) spawnable distros in staging, and also verify that the instance type is available. * ap-southeast-2 * Europe (Ireland): eu-west-1 * California: us-west-1 ""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12387","06/25/2020 16:16:17",1,"Add golint for jasper ""Jasper tends to be well-documented and it would be good to ensure future devs continue keeping the documentation in a good state.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12388","06/25/2020 16:49:57",1,"Do not set subprocess.scripting working directory in meta-build ""Since it should be set by default to the task's working directory.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12389","06/25/2020 17:39:26",2,"Configure host events ""Set up AWS to send webhooks for these [events|https://docs.aws.amazon.com/AmazonCloudWatch/latest/events/EventTypes.html#spot-instance-event-types]: * Spot Instance Interruption and/or termination * instance running # Create a SNS topic and add a https subscription to send webhooks to evergreen (separate topics/subscriptions for prod/staging?) # Configure a CloudWatch/EventBridge rule to publish to the topic""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12390","06/25/2020 17:46:10",2,"Add REST route to handle AWS webhooks ""Probably one route will suffice""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12391","06/25/2020 17:47:49",2,"Terminate Spot Instances that are about to terminate ""For the spot instance termination event we should terminate the host instead of waiting for AWS to kill it (they aren't running tasks anymore anyway).""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12392","06/25/2020 17:48:50",2,"Mark terminated instances as terminated in the DB ""If AWS tells us a host has terminated we should mark it in the db if we haven't already.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12393","06/25/2020 17:51:40",2,"Transition running hosts to running ""When we find out a userdata host is running we should skip the provisioning state (and the setupHostJob) and just go straight to running. We'll need to relocate OnUp and setDNSName, maybe to cloud-host-ready. This will allow us to remove the logic in next_task that lets provisioning hosts run tasks. Also, we could remove the user-data-done job since the agent won't be asking for tasks until userdata has finished running. ""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12396","06/25/2020 21:45:37",2,"Open Page in the New UI displays on pages that are not yet supported ""Presently, clicking this link on a [legacy UI commit page|https://evergreen.mongodb.com/version/mongodb_mongo_master_47dcbeba765b04a829be9c27289e152e24e1a855] results in a GraphQL error. Clicking it on the [task details page|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_dynamic_required_lint_eslint_47dcbeba765b04a829be9c27289e152e24e1a855_20_06_25_20_10_38] shows a page in the new UI, as if it were a patch, but then clicking on the Patch breadcrumb leads to the same GraphQL error. Removing the Open Page in the New UI link on the commit details/task details pages would avoid potential confusion about which pages ought to be supported. It might also be worthwhile to consider displaying a more specific error message in Spruce when a user tries to navigate to a page that isn't yet supported.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12399","06/26/2020 17:33:23",1,"Failures: test-rest-model on Ubuntu 16.04 [Evergreen @ 18ab08f5] (TestCodegen, aliases, basic, object_embedded, object_simple, scalars, thirdparty) "" h2. [test-rest-model failed on Ubuntu 16.04|https://evergreen.mongodb.com/task/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47/0] Host: [ec2-54-91-138-40.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-08c55209d62321a1b] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/18ab08f5613814fd6d8dc97de4c47c6db7d08f11]: EVG-8169 allow aliasing fields (#3717) | 25 Jun 20 19:46 UTC Evergreen Subscription: ; Evergreen Event: *TestCodegen* - [Logs|https://evergreen.mongodb.com/test_log/5ef4ffe5a4cf4714d36fb8c2] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47#TestCodegen=fail] *aliases* - [Logs|https://evergreen.mongodb.com/test_log/5ef4ffe5a4cf4714d36fb8c2] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47#aliases=fail] *basic* - [Logs|https://evergreen.mongodb.com/test_log/5ef4ffe5a4cf4714d36fb8c2] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47#basic=fail] *object_embedded* - [Logs|https://evergreen.mongodb.com/test_log/5ef4ffe5a4cf4714d36fb8c2] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47#object_embedded=fail] *object_simple* - [Logs|https://evergreen.mongodb.com/test_log/5ef4ffe5a4cf4714d36fb8c2] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47#object_simple=fail] *scalars* - [Logs|https://evergreen.mongodb.com/test_log/5ef4ffe5a4cf4714d36fb8c2] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47#scalars=fail] *thirdparty* - [Logs|https://evergreen.mongodb.com/test_log/5ef4ffe5a4cf4714d36fb8c2] | [History|https://evergreen.mongodb.com/task_history/evergreen/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47#thirdparty=fail] [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_ubuntu1604_test_rest_model_18ab08f5613814fd6d8dc97de4c47c6db7d08f11_20_06_25_19_46_47/0?type=T] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-12400","06/26/2020 18:33:28",2,"No error is displayed when modifying a host that is terminated ""When using the CLI to extend the expiration time of a host using evergreen host modify --host --extend 5 with an id of a host that is terminated, the CLI returns """"Successfully queued changes to spawn host with ID 'hostID'.""""""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12404","06/29/2020 17:58:36",1,"Hosts that fail to convert provisioning type should be set to running ""OSX hosts failed to convert to new provisioning due to a permission denied (presumably because the sudo permissions to allow new provisioning were only applied to macos-1012 and macos-1014, but not osx-1010). However, if you switch back to legacy provisioning, the host is not set to running automatically.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12412","06/29/2020 21:03:02",2,"ansi_to_html library sometimes errors when viewing logs ""the error happens when viewing these pages [https://spruce.mongodb.com/task/sqlproxy_full_matrix__os_full_matrix~osx_mongodb_version~latest_mongodb_topology~standalone_module_tests_patch_129f80fcf3b75ecb4f8f24ecbb1ba2b5d7dc1a3e_5efa177e3e8e86592c495ab0_20_06_29_16_32_00/logs] [|https://spruce.mongodb.com/task/buildhost_configuration_ubi7_build_container_patch_3899b4ba54eb93ecd70b2963ae6c380b83ee4f41_5f05fc4c1e2d17015edcdc21_20_07_08_17_03_08/logs] [https://app.bugsnag.com/mongodb/evergreen/errors/5efa1a85442e1f001893f433?filters[event.since][0]=30d&filters[error.status][0]=new]""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12413","06/29/2020 22:13:31",1,"Shrub should support yaml output ""Shrub currently outputs a JSON Evergreen configuration file. While this is fine for things like the agent which just parses it, it's not a very human-readable format. Since we will eventually support local task generation for patches, it would be nice if it had an option to output YAML instead of JSON.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12414","06/29/2020 22:16:17",2,"Move Jasper metabuild to Evergreen ""The metabuild tool should belong in evergreen since it's an Evergreen-focused feature. It would be good to keep features related to evergreen configuration within Evergreen rather than put it in a repository (which could someday integrate with it, but currently doesn't and likely won't in the near or medium term).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12418","06/30/2020 15:25:05",2,"Versions with blocked tasks aren't marked as complete ""We're supposed to mark a version as complete when all its remaining tasks are blocked, but this isn't happening.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12419","06/30/2020 15:52:23",1,"Fix leaking resources for TestLogMergeIterator ""Seems like there is a leaked resource in the TestLogMergeIterator test, the tmp dir cannot be removed at the end of the test. It only fails on Windows. ""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12422","06/30/2020 17:15:49",2,"Error restarting failed tasks in patch build with ""duplicate key error"" ""In [this patch|https://spruce.mongodb.com/version/5efa3c6ec9ec4470c51e2319/tasks], if I """"Restart"""" all tasks that failed with """"System Failure"""", I can reliably get the following error: ""","Error while restarting patch: 'GraphQL error: error restarting patch: failed to archive task: task.Archive() failed: multiple write errors: [{write errors: [{E11000 duplicate key error collection: mci.old_tasks index: _id_ dup key: { : """"mongodb_mongo_master_enterprise_rhel_62_64_bit_dynamic_required_initial_sync_fuzzer_1_enterprise_rhel_62_64_bit_dynamic_required_patch_847e085c0fb7621..."""" }}]}, {}]' ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12427","06/30/2020 20:40:38",2,"LogIterator implementations do not close readers ""The LogIterators call {{(pail.Bucket).Get()}} but do not have a way to close the returned manually {{io.ReadCloser}} (e.g. after the {{LogIterator}} is done). This is causing the Windows tests to fail with device/resource busy when trying to clean up the local pail bucket (TestMergeLogIterator). One possible solution would be to provide a {{Close}} method in the {{LogIterator}} interface.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12432","07/01/2020 15:27:45",1,"Create protobuf for system metrics service ""This ticket requires you to create the protocol buffer (gRPC) for the system metrics collection service. There are some examples you can look at including [buildlogger|[https://github.com/evergreen-ci/cedar/blob/master/buildlogger.proto]] and [perf|[https://github.com/evergreen-ci/cedar/blob/master/perf.proto]]. Think about what users would want to send and how that fits with the functions you have written in the model package.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12440","07/01/2020 19:37:10",1,"Implement AppendArtifact on the system metrics model ""This should upload a new artifact to the bucket specified in the SystemMetricsArtifactInfo struct and append the key to the slice of keys.""","",1,0,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12442","07/01/2020 21:17:15",1,"Cedar CLI client commands are not usable ""It requires a host/port from the parent CLI context, which does not have those flags defined. It should just get the flags from the CLI context at the client level. It also requests from the wrong (because it needs to be prefixed with {{/rest/v1}}.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12452","07/07/2020 14:26:16",0,"Add short flags for revendor command ""Mostly so that I can type less.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12453","07/07/2020 15:48:59",2,"Create API to test a project alias ""It can mostly be a wrapper around https://github.com/evergreen-ci/evergreen/blob/c19b96fbffa64becbfabc7e1bbe13e0425b9475a/model/project.go#L1282""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12539","07/08/2020 12:38:39",1,"Create new evergreen project for Spark connector ""Hi, Could I get another new evergreen project for the Spark Connector? A copy of https://evergreen.mongodb.com/waterfall/mongo-spark-2.3.x would be great, just need it to point to the: 2.4.x branch. Many thanks, Ross""","",0,0,0,1,0,0,0,0,0,0,0,0,0,0,1
+"EVG-12464","07/08/2020 14:16:31",2,"UI should handle when patch has no base commit ""This will happen rarely, for example in a new project that has no commits yet""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12467","07/08/2020 19:24:58",2,"CLI should support SendMessages ""This is so the CLI maintains parity with the other remote interfaces and so the the CLI over SSH client can call {{SendMessages}}.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12470","07/09/2020 08:37:06",1,"Enable BB plugin for ""wiredtiger-mongo-v4.4"" Evergreen project ""Please help to enable BB plugin for the [wiredtiger-mongo-v4.4|https://evergreen.mongodb.com/waterfall/wiredtiger-mongo-v4.4] project. The plugin had been enabled for the other 2 wiredtiger projects previously if that helps. Cheers.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,0,1
+"EVG-12471","07/09/2020 12:44:55",2,"Add authentication support to evergreen project subscription POST configuration ""As part of the PM-1800 / BF Matching for Patch Builds, DAG would like the build baron server to receive patch failure notifications. Since this project is implemented as a kanopy service, it is behind [corp secure|https://wiki.corp.mongodb.com/display/SEC/CorpSecure+2.0]. We will be requesting a long lived JWT to grant access to our kanopy service but evergreen will need to support JWT (as a secure field) in order to post to the /patch_build/add_instance endpoint. A JWT can be used either as a [cookie or an authorization header|https://jira.mongodb.org/browse/DAG-274?focusedCommentId=3276831&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#comment-3276831].""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12477","07/09/2020 18:05:40",1,"Shrub should omitempty optional fields ""This will cut down on the verbosity of the generated configs.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12478","07/09/2020 18:10:50",2,"Versions triggered from git tags shouldn't be set inactive ""These should always be scheduled.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12487","07/10/2020 16:26:08",1,"e2e tests: fix task_actions spec ""these tests flake: Clicking on Restart and then Schedule should produce an error banner There should be 2 banners visible from the previous actions ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12497","07/10/2020 19:44:39",1,"Create new evergreen project for mongo-java-driver-4.1 ""Please create new Evergreen project: Name: mongo-java-driver-4.1 Display Name: Java Driver (4.1) Branch Name: 4.1.x Otherwise, it should be a clone of mongo-java-driver-4.0, with all the same project properties, etc. Please also create the necessary S3 buckets to allow uploads of build artifacts ""","",0,0,0,1,0,0,0,0,0,0,0,0,0,0,1
+"EVG-12498","07/10/2020 19:50:40",1,"Create mongo-go-driver-v1.4 Evergreen project ""Please create a new Evergreen project: Name: mongo-go-driver-v1.4 Display Name: Mongo Go Driver (release/1.4) Branch Name: release/1.4 Otherwise, it should be a clone of mongo-go-driver, with all the same project properties, etc. Please also create the necessary S3 buckets to allow uploads of build artifacts""","",0,0,0,1,0,0,0,0,0,0,0,0,0,0,1
+"EVG-12491","07/11/2020 23:35:56",1,"Tasks failed to restart ""See https://mongodb.slack.com/archives/G979SL0HM/p1594502408091800 for discussion. A system-failed task failed to restart. Reverting task group reset logic commits fixed it. Tasks that failed to restart before the revert: * https://evergreen.mongodb.com/task/mongo_release_rhel80_compile_2dae2754daded5fb71a93d19223613850826f75e_20_07_10_19_05_28 * https://evergreen.mongodb.com/task/mongo_release_amazon2_compile_2dae2754daded5fb71a93d19223613850826f75e_20_07_10_19_05_28 Commits I reverted: * 1c22aa164c4c608a26103d969b4ccc51253595f8 * 20482f76c62c0971e55051de639ce32c48608573 I reverted the first because just reverting the second got me into some pretty unpleasant merge conflicts.""","",0,0,0,0,0,0,0,1,0,1,0,0,0,0,0
+"EVG-12493","07/13/2020 15:33:57",2,"Remove dependence on working directory for clone ""The working directory flag for Make-based projects should be removed from the CLI. Users should be able to specify the working directory to clone to, but it should just default to the task's working directory (i.e. {{${workdir}}}). For golang-based projects, the working directory should be configurable just like the Make one, but it should default to {{${GOPATH}/src/}} if none is specified.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12495","07/13/2020 16:17:47",2,"GQL query for Host ""create a Host gql type that has all of the host metadata shown on these pages then create a host query to fetch a host by it's id only include recentEvents in this structure if it's not meant to be paginated [https://mongodb.invisionapp.com/d/main/#/console/19647524/424885179/comments] [https://mongodb.invisionapp.com/d/main/#/console/19647524/424885176/comments] ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12496","07/13/2020 16:20:11",2,"GQL query for myHosts ""myHosts gql query should return a list of hosts that will support this page. since a user can only have 6 hosts, it's unlikely we will need to paginate this query [https://mongodb.invisionapp.com/d/main/#/console/19647524/424885179/comments]""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12502","07/13/2020 17:40:04",1,"Allow user to specify generated config output format in CLI ""Since shrub now supports JSON and YAML, we should generate configs for either one (the YAML one would be better for human readability while the JSON one would be suitable for {{generate.tasks}}).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12503","07/13/2020 19:19:35",2,"Task checkboxes remain unselected after clicking ""On some variants, the task check boxes appear to be unselected even after clicking individually or clicking """"Select all tasks in this variant"""". This was the patch build: https://spruce.mongodb.com/patch/5f0ca3d7306615237fb06b7b/configure/tasks I observed this on the """"OCSP tests"""" variant. Reproduced in Google Chrome 83.0.4103.116. Screenshot attached.""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12513","07/14/2020 15:30:12",2,"Host Metadata Card Component ""Render the metadata card component: query the data and display it. !image-2020-07-14-10-43-52-718.png!""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12514","07/14/2020 15:39:15",2,"Create Route ""Create the /host/ route ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12515","07/14/2020 15:39:54",2,"Create page layout ""Create the /host/ page layout ""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12517","07/14/2020 15:43:16",2,"Display title and status ""Display the title and status for the /host/ page !image-2020-07-14-10-43-04-831.png!""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12523","07/14/2020 16:08:50",2,"Restart Jasper button ""Button that launches modal and can be used on /hosts and /host pages""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12524","07/14/2020 16:09:53",2,"Selecting multiple hosts on hosts page: mixed state checkbox ""Mixed state checkbox in column header""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12527","07/14/2020 16:11:54",2,"Reusable table filters ""Clicking `reset` removes query param from url""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12528","07/14/2020 16:12:16",2,"Filtering hosts: dropdowns UI `search/filter` button ""Clicking `search/filter` adds query param to url""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12529","07/14/2020 16:12:38",2,"Filtering hosts: fetch filtered hosts ""Listen to changes in url params and query hosts with params as filters""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12530","07/14/2020 16:13:08",2,"Hosts page pagination ""* Get next page of results * Update ‘page’ url param & re-fetch * Choose number of hosts per page * Update `limit` url param & re-fetch""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12533","07/14/2020 19:15:57",1,"host configure commands appear to hang ""printing to stdout stops after the first command""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12535","07/15/2020 15:14:25",2,"Set read concern majority ""We set write concern majority. We should also set read concern majority to avoid reading data that can be rolled back. Note that the environment variable that sets write concern is in 10gen/kernel-tools. We should use a similar model for setting read concern: Support setting arbitrary read concerns in the app, with a default, and pass read concern majority in the env.""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12537","07/15/2020 18:45:36",1,"Disable oom tracker in containers ""On linux, the container won't have access to the host's dmesg buffer and trying to run it will return """"dmesg: read kernel buffer failed: Operation not permitted"""" https://stackoverflow.com/questions/41171349/why-is-a-shell-within-a-docker-container-showing-dmesg-content-from-the-host""","",0,0,1,0,0,0,0,0,0,1,0,0,1,0,0
+"EVG-12538","07/15/2020 19:22:56",1,"Have timber buildlogger client take username and apikey ""This will replace the current way of asking for the certificate files.""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12540","07/15/2020 20:07:24",2,"Allow FTDC to collect JSON in memory ""FTDC currently supports collecting JSON and writing the FTDC chunk to a file. However, it may be useful to keep it in memory.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12544","07/15/2020 23:05:46",2,"OOM Tracker misses pids ""There are times when the OOM tracker misses pids that were OOM killed [One example|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_majority_read_concern_off_replica_sets_large_txns_format_2_enterprise_rhel_62_64_bit_majority_read_concern_off_74ab0cdac56e1cc62fc96d3ca3be3ddfa54f2bcb_20_07_11_23_55_48#/log/S] where we found 12173 but missed 16811 (which is included in the system log) [Another example|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_majority_read_concern_off_replica_sets_large_txns_format_2_enterprise_rhel_62_64_bit_majority_read_concern_off_74ab0cdac56e1cc62fc96d3ca3be3ddfa54f2bcb_20_07_11_23_55_48#/log/S], and [another|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_dynamic_required_gle_auth_patch_188c858a12705164b7ae5f1ad418821bedd94bb4_5f0caad39ccd4e2f8b9748c4_20_07_13_18_41_44#/log/S#comparehashes=188c858a12705164b7ae5f1ad418821bedd94bb4&threads=maxonly]""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12545","07/15/2020 23:06:47",2,"Display task with an OOM kill? ""The [task|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_majority_read_concern_off_display_replica_sets_large_txns_format_patch_c957403d177c8b83c3d3693b29d7af73f34ffdfc_5f0c24f0e3c3311dad2f5e25_20_07_13_09_10_09#/log/S#comparehashes=c957403d177c8b83c3d3693b29d7af73f34ffdfc&threads=maxonly] in question""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12546","07/15/2020 23:34:58",2,"Detected OOM Kill when there was none in the system logs ""This [task|https://evergreen.mongodb.com/task/mongodb_mongo_master_enterprise_rhel_62_64_bit_dynamic_required_sharding_ese_gcm_37_enterprise_rhel_62_64_bit_dynamic_required_patch_188c858a12705164b7ae5f1ad418821bedd94bb4_5f0caad39ccd4e2f8b9748c4_20_07_13_18_41_44#/log/S#comparehashes=188c858a12705164b7ae5f1ad418821bedd94bb4&threads=maxonly] detected an OOM kill with no PIDs. The system logs don't mention an OOM Kill""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12548","07/16/2020 16:16:59",1,"Remove restriction of hosts greater than half the number of task group tasks ""There is a restriction on task groups that the max number of hosts must not be more than half the number of tasks. I believe this was meant to enforce that people use task groups to serialize tasks. However, this doesn't really make sense. Users might want the task group behavior of preserving the directory but also be fine with parallelism.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12554","07/16/2020 21:47:35",2,"Task has nonexistent dependency ""https://evergreen.mongodb.com/task/mongo_python_driver_Coverage_Report_coverage_report_patch_5d92b2f5523876653018a4427962ef8ee8f7e6ff_5f0638a6d1fe074eae3090f0_20_07_08_21_20_39 It has a very long list of dependencies that are not specified in the yml. Also it mysteriously got scheduled a week after the patch was submitted""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12555","07/17/2020 13:30:04",1,"Increase timeout for Task S3 downloads to 30 minutes ""In order to support slower network speeds on static hosts (Z,PPC), we want to increase the timeout on downloading artifacts from S3. From manual tests, 30 minutes seems to be a safe threshold. Example: https://evergreen.mongodb.com/task/mongodb_mongo_v4.4_enterprise_rhel_81_ppc64le_failpoints_31b08e7d7e095132b7db3e8f4a4271478cf353ab_20_07_16_02_59_08 ""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12558","07/17/2020 14:58:55",2,"Prefrences page: GQL Mutation to update a public key ""Prefrences page: GQL Mutation to update a public key""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12557","07/17/2020 14:58:55",2,"Preferences page: myPublicKeys query ""create a query to request the authenticated users public keys. ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12556","07/17/2020 14:58:55",2,"Prefrences page: GQL Mutation to create a public key for the authenticated user ""Prefrences page: GQL Mutation to create a public key for the authenticated user""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12562","07/17/2020 14:58:56",2,"Prefrences page: Create the ""+ Add new key"" button on Manage Public Keys page ""Prefrences page: Create the """"+ Add new key"""" button on Manage Public Keys page as well as the component (needs design) to submit a new key. Hook up the component submit button to the createPublicKey mutation.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12561","07/17/2020 14:58:56",2,"Prefrences page: GQL Mutation to delete a public key for the authenticated user ""Prefrences page: GQL Mutation to delete a public key for the authenticated user""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12560","07/17/2020 14:58:56",2,"Prefrences page: Render Key Name and Action buttons in the Manage Public Keys page ""Prefrences page: Render Key Name and Action buttons in the Manage Public Keys page. Hook up trash can button to the delete public key mutation.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12559","07/17/2020 14:58:56",1,"Prefrences page: Add Manage Public Keys tab in the Preferences page and a placeholder page ""Prefrences page: Add Manage Public Keys tab in the Preferences page and a placeholder page""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12564","07/17/2020 14:58:57",2,"Prefrences page: Hook up the pencil action button in the Public Keys table ""Prefrences page: Hook up the pencil action button in the Public Keys table to edit the public key. Clicking on the key should open a component to edit the key (needs design)""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12563","07/17/2020 14:58:57",1,"Spawn host page: Page title + badges ""Spawn host page: Page title + badges""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12568","07/17/2020 14:58:58",2,"Spawn host page: GQL mutation to start a paused host ""Spawn host page: GQL mutation to start a paused host""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12567","07/17/2020 14:58:58",2,"Spawn host page: GQL mutatation to delete host ""Spawn host page: GQL mutatation to delete host""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12566","07/17/2020 14:58:58",2,"Spawn Host page: GQL mutation to pause/restart host ""Spawn Host page: GQL mutation to pause/restart host""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12570","07/17/2020 14:58:59",2,"Spawn host page: Create GQL mutation to spawn a host ""Spawn host page: Create GQL mutation to spawn a host""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12569","07/17/2020 14:58:59",8,"Spawn host page: Create spawn host modal and hook up to spawn host button ""Spawn host page: Create spawn host modal and hook up to spawnHost mutation and spawn host button""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12573","07/17/2020 14:59:00",2,"Spawn host page: GQL query to get spawnableDistros ""Spawn host page: GQL query to get spawnableDistros""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12572","07/17/2020 14:59:00",2,"Spawn host page: GQL mutation to edit a spawned host ""Spawn host page: GQL mutation to edit a spawned host""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12576","07/17/2020 14:59:01",2,"Spawn host page: Render table rows ""Spawn host page: Render table rows ""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12575","07/17/2020 14:59:01",2,"Spawn host page: Create table dropdown card. Populate with data from host query. ""Spawn host page: Create table dropdown card. Populate with data from host query.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12574","07/17/2020 14:59:01",2,"Spawn host page: GQL query to get instanceTypes ""Spawn host page: GQL query to get instanceTypes""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12579","07/17/2020 14:59:02",2,"Volumes page: GQL mutation to unmount / mount volume from host ""Volumes page: GQL mutation to unmount / mount volume from host""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12578","07/17/2020 14:59:02",2,"Volumes page: GQL mutation to delete volume ""Volumes page: GQL mutation to delete volume""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12577","07/17/2020 14:59:02",1,"Volumes/Host page: Create tab for Volumes and Hosts and empty placeholder pages ""Volumes/Host page: Create tab for Volumes and Hosts at /spawn route and empty placeholder pages""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12582","07/17/2020 14:59:03",2,"Volumes page: Create table dropdown card component - hook up to volumes query ""Volumes page: Create table dropdown card component - hook up to volumes query""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12580","07/17/2020 14:59:03",2,"Volumes page: GQL query for myVolumes ""Volumes page: GQL query for myVolumes, include these fields in the Volumes gql type: volume, mountedOn, status, expiresAt, createdAt, startedAt (maybe), type, size, availabilityZone, isHomeDir""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12584","07/17/2020 14:59:04",2,"Volumes page: GQL mutation to create volume ""Volumes page: GQL mutation to create volume""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12583","07/17/2020 14:59:04",2,"Volumes page: GQL mutation to edit volume ""Volumes page: GQL mutation to edit volume""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12588","07/17/2020 14:59:06",1,"Volumes page: Page title + badges ""Volumes page: Page title + badges""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12603","07/20/2020 18:39:09",2,"Suggest re-logging into UI if evergreen CLI cannot authenticate user ""User refresh tokens expire after 100 days and the only way to get a new token is to log into Okta via the UI. However, the CLI currently just returns 401 unauthorized - it should suggest logging in again if it gets this due to expired tokens. The way the middleware is written, there's no way to distinguish between different reasons for 401 unauthorized (e.g. unauthorized because ) since the user checking is based on the presence/absence of a {{gimlet.User}} being attached, so it's likely that suggesting logging in again will be the wrong solution since it's just a guess.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12605","07/20/2020 19:33:24",2,"Generated task ran 8 hours after last task it depended on finished ""This task ran on Jul 17 at 1:49am [https://evergreen.mongodb.com/task/mms_e2e_local_with_daemon_foliage_dd9fa2b4001cd68b7d427b3ff9e6e8924af9b5fb_20_07_16_13_59_36#/log/EV] It ran over 8 hours after the last task it depended on finished on Jul 16 at 5:05pm [https://evergreen.mongodb.com/task/mms_e2e_local_with_daemon_E2E_Local_BRS_Automated_PIT_Restore_ShardedCluster_Blockstore_S3_dd9fa2b4001cd68b7d427b3ff9e6e8924af9b5fb_20_07_16_13_59_36] This was on rhel70-large over the evening which we would expect to not be constrained on hosts, so we're perplexed what caused the delay The task was generated long before on Jul 16 at 10:12am [https://evergreen.mongodb.com/task/mms_evg_foliage_stat_check_init_dd9fa2b4001cd68b7d427b3ff9e6e8924af9b5fb_20_07_16_13_59_36] This gave us a makespan of 15 hours for the variant. We need to bring that down. Looking for a root cause of why this 8 hour delay occurred, any steps the evergreen team or we can implement to prevent it in the future. CC: [~geoff.mishkin] [~james.broadhead]""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12609","07/20/2020 20:41:01",2,"Task restarted with system failure, then marked started twice ""[https://evergreen.mongodb.com/task/mms_e2e_local_with_daemon_E2E_Local_BRS_Automated_PIT_Restore_ShardedCluster_Blockstore_S3_dd9fa2b4001cd68b7d427b3ff9e6e8924af9b5fb_20_07_16_13_59_36#/log/EV] We need to know if this behavior is something expected we need to change batch times or test lengths to account for, in order to make a task's contribution to the variant makespan shorter. Or if it is something that needs to be addressed by the evergreen team to prevent from happening again.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12613","07/20/2020 23:25:02",2,"Hosts page table ""[https://mongodb.invisionapp.com/share/C2WKQWRWG5Y#/screens/424885145]""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12624","07/21/2020 19:45:02",2,"Are task groups clearing the work dir between tasks within the group? ""https://github.com/evergreen-ci/evergreen/blob/2a733770e298ce01069e5c916b6a4d8773518c3d/agent/agent.go#L323 doesn't seem to check, but also I think we would be hearing about it if this were the case. To test this, you can make a task group that writes a file in the first task and then reads that file in the second task.""","",0,0,0,0,0,1,0,0,0,1,0,0,0,0,0
+"EVG-12625","07/21/2020 19:54:00",1,"Rename CreateSystemMetricRecord to CreateSystemMetricsRecord and fix panic logging in timber ""The naming change is just for consistency with other functions, which use the plural form.""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12626","07/21/2020 21:53:15",2,"LoggingCache does not work as intended ""It should be a logging cache stored on the remote service side, not the client side. It should probably follow the REST model for the logging cache implementation, where it connects to the REST service and makes requests for all the logging cache methods. We should test the logging cache once it's implemented (along with {{SendMessages}}, which depends on this).""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12629","07/22/2020 14:00:37",2,"Handle expired refresh tokens by logging out user ""If the user has an expired refresh token, the background reauthentication job should clear their login cache as soon as it detects that error so that they must log in again. Gimlet will have to return a special error message for the case of a user's refresh token being expired.""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12632","07/22/2020 20:59:36",1,"`evergreen admin enable-service` and `disable-service` fetch blank service flags ""They make a request to the legacy admin settings route, which returns an empty struct (except for the banner). We should fix this so it actually returns the existing service flags. Otherwise, these CLI commands will unset all the service flags.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12640","07/23/2020 19:44:00",1,"Add exit status to cedar GRPC client ""Just in case the metrics collector fails.""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12642","07/23/2020 21:07:51",2,"Don't stop starting a host if can't get price information ""EVG-12641 might be a situation where hosts can actually start, but the pricing API blocks Evergreen from setting hosts to running. We should put a context on calls to the cachingPriceFetcher and not error if they fail, since we can probably safely continue, since we no longer rely on it to get the cheapest AZ.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12644","07/23/2020 21:56:56",1,"Create script to easily convert mongo documents to json ""This would help us quickly turn prod mongodb document data into plain json so that it can be used to seed the local data This would greatly speed up the process of generating data to seed the local db.""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12648","07/24/2020 14:50:19",2,"Initial task setup failure for task ""This is not related to problem creating certificates and occurred less than 1% of the time. https://evergreen.mongodb.com/task/mongodb_mongo_master_amazon_compile_78cc027285359cc21c55f2dac69ce41cb9f23959_20_07_24_10_05_58""","",0,0,0,0,0,0,0,1,0,1,0,0,0,0,0
+"EVG-12649","07/24/2020 19:15:35",1,"Fix make dist to compile for the target GOOS operating system ""This has the same problem that Barque had where it'll compile for the native computer's GOOS instead of the one set in the command line (e.g. {{GOOS=linux make dist}})""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12670","07/27/2020 20:20:33",2,"Implement LTS/continuous/development checks for new versioning scheme ""The new versioning implementation should distinguish between LTS releases and non-LTS releases (i.e. continuous and development releases).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12679","07/28/2020 14:46:59",1,"Remove freebsd from cross-compilation ""Nobody uses it anymore so we can get rid of it.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12681","07/28/2020 15:59:20",1,"Can't patch with apple's version of git ""The version string for Apple's version of git doesn't match the [regexp|https://github.com/evergreen-ci/evergreen/blob/4dab3e36c27e0487aa3cf2c8c7da37e1c0d29259/operations/patch_util.go#L673] ""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12682","07/28/2020 18:20:38",8,"Test new versioning scheme end to end ""We should verify that we can correctly put a version greater than 4.5.0 into the development repository. It may also make sense to do this with a release candidate. This has to be done before we can do the first continuous release.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12684","07/28/2020 19:55:18",2,"Fleet errors for a set of distros ""We're getting many errors trying to spawn ubuntu*-arm64-large hosts.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12686","07/28/2020 20:02:55",0,"Remove nonce error checking for ID token validation ""After filing [this issue|https://github.com/okta/okta-jwt-verifier-golang/issues/43], it seems like it should no longer be necessary to check for an error string related to empty nonce claim when returned an ID token [here|https://github.com/evergreen-ci/gimlet/blob/d28831ee71d1d02871f7e6a1c94ec5387430218d/okta/okta.go#L196-L205]. We should be able to remove this check.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12687","07/28/2020 21:13:11",1,"Can't create empty patches ""An empty patch errors: {code:bash} $ evergreen patch -p evergreen Enter a description for this patch (optional): Unexpected reply from server (500 Internal Server Error): 400 (Bad Request): error getting summaries by commit: error getting patch summaries for commit 'master: EVG-12681 fix for Apple's git (#3834)': couldn't validate patch: Error waiting on patch: 562 - fatal: unrecognized input : exit status 128 {code} """," $ evergreen patch -p evergreen Enter a description for this patch (optional): Unexpected reply from server (500 Internal Server Error): 400 (Bad Request): error getting summaries by commit: error getting patch summaries for commit 'master: EVG-12681 fix for Apple's git (#3834)': couldn't validate patch: Error waiting on patch: 562 - fatal: unrecognized input : exit status 128 ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12690","07/29/2020 16:44:40",8,"Implement new versioning scheme ""This should ideally be a separate implementation from the legacy versioning scheme, whose logic should preferably not change. For this, we should just implement the existing methods on the mongodb version object, but with their new meanings. This new versioning scheme should only apply for versions 4.5.0 and above. For example, development series do not exist anymore, but a release that goes into the development repo should be denoted by any non-LTS version which has {{major.quarter.patch}} format and is not followed by {{-rcN}} (special for release candidates).""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12696","07/29/2020 23:54:55",2,"New commits are not being tracked by mongo-go-driver-v1.4 ""The mongo-go-driver-v1.4 project isn't automatically scheduling waterfall builds for new commits to the branch it's tracking. [~john.liu] initially fixed this by using the """"force run repotracker on save"""" option in the project settings, which kicked off one waterfall run, but the project did not pick up commits the next day either.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12697","07/30/2020 00:53:05",1,"Call update function onEnter for table filters ""In table dropdown filters, the update function should be called if user presses Enter""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12712","07/31/2020 19:36:48",1,"Allow users to copy-to-clipboard build variant names on patch configuration page ""This is useful if a user wants to copy/paste build variant names for documentation, conversations about which task(s) to schedule, etc. Presently, it is not possible to select them (see this [patch build|https://spruce.mongodb.com/patch/5f244d89d6d80a20695854b1/configure/tasks]). ""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12717","08/03/2020 17:22:57",2,"Evergreen automatically scheduled a zSeries task from March 24th 2020 on August 2nd 2020 ""Tasks older than 7 days shouldn't be getting automatically scheduled. https://evergreen.mongodb.com/task/mongodb_mongo_v4.4_enterprise_rhel_72_s390x_inmem_sharding_op_query_1_enterprise_rhel_72_s390x_inmem_70c19c844d348774290fdadb9f9bbb0deb5301ad_20_03_24_06_16_46#/log/EV |Aug 2, 2020 6:45:04 am|Scheduled at Aug 2, 2020, 6:45:03 am| |Aug 2, 2020 7:00:31 am|Dispatched to host [rhel74-z-12.dallasisv.build.10gen.cc|https://evergreen.mongodb.com/host/rhel74-z-12.dallasisv.build.10gen.cc]| |Aug 2, 2020 7:00:37 am|Marked as """"started""""|""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12724","08/03/2020 23:18:23",2,"Tasks submitted time should be their ingest_time ""The submitted time shown for a task taken from the task's create_time field. Given that a patch task's create time is [that of its base version|https://github.com/evergreen-ci/evergreen/blob/d40dbd00d349328491003a6fc21bc28fae5c8ef6/model/lifecycle.go#L1040] this is confusing. It would be better to show the task's ingest_time.""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12729","08/05/2020 17:55:48",2,"Implement Download method ""This should download a single metric type from s3 for a SystemMetrics object.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12730","08/05/2020 19:42:00",2,"Host Events Query ""Fetch host event data on the ui. ""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12736","08/06/2020 16:58:47",2,"Upgrade gopsutil to latest tagged release ""The version of gopsutil that we're depending on panics on MacOS when calling {{(*Process).CreateTime()}} due to [a bug that has been fixed in a later version|https://github.com/shirou/gopsutil/issues/799#issuecomment-562764688]. This will require some work to fix grip, which is pinned to an old version of gopsutil.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12737","08/06/2020 17:21:41",2,"Can you create a patch against a restricted project you don't have access to? ""[https://mongodb.slack.com/archives/C0V896UV8/p1596729650420200] Initial thoughts are yes, but I don't have a restricted project that I don't have access to for testing. If you can, then we should probably change this issue to a bug, since you could leak Evergreen project variable values in the patch.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12738","08/06/2020 17:56:21",0,"Handle nil time and duration in jasper RPC conversions ""{{ptypes.Timestamp}}/{{ptypes.Duration}} will fail if passed a nil {{time.Time}}/{{time.Duration}} (e.g. if the process hasn't finished yet so doesn't have an ending timestamp)""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12744","08/07/2020 19:21:42",3,"Create global_max_hosts functionality ""Currently, we are relying on max_host limits to prevent the total number of hosts in the system from becoming too great. We should have a global limit in place so that we can loosen up max_host limits on individual distros, which are frequently pegged at max_hosts and build up very long wait times as a consequence. use this data: https://mongodb.splunkcloud.com/en-US/app/search/search?earliest=rt-1m&latest=rt&q=search%20index%3Devergreen%20%22host%20stats%20by%20distro%22&display.page.search.mode=fast&dispatch.sample_ratio=1&display.general.type=events&sid=rt_1596821781.155325 Pegged hosts: https://mongodb.splunkcloud.com/en-US/app/search/search?earliest=-60m%40m&latest=now&q=search%20index%3Devergreen%20%22host%20stats%20by%20distro%22%20%7C%20spath%20path%3Ddata%7B%7D%20output%3Ddata%20%7C%20fields%20-%20_raw%20%7C%20fields%20data%20%7C%20mvexpand%20data%20%7C%20spath%20input%3Ddata%20%7C%20where%20status%3D%22running%22%20AND%20provider!%3D%22static%22%20and%20distro!%3D%22ubuntu1804-workstation-cluster%22%20and%20distro!%3D%22ubuntu1804-workstation%22%20%7C%20stats%20avg(max_hosts)%20as%20_max_hosts%20avg(count)%20as%20avg_count%20by%20distro%20%7C%20eval%20pegged_floor%3Dfloor(_max_hosts%20*%200.99)%20%7C%20where%20avg_count%20%3E%20pegged_floor%20%7C%20sort%20by%20pegged_floor&display.page.search.mode=verbose&dispatch.sample_ratio=1&display.general.type=visualizations&display.page.search.tab=visualizations&display.visualizations.charting.chart=column&sid=1596824255.157079 ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12746","08/08/2020 10:24:33",1,"The `evergreen` command line utility fails on Windows due to format of Git version ""The `evergreen` command line utility for Windows fails with this error: The version that Git reports on Windows indeed is """"2.15.0.windows.1"""": """," E:\workspace\mongo>evergreen patch -d """"SERVER-49192"""" Uncommitted changes are omitted from patches by default. Use the '--uncommitted, -u' flag or set 'patch_uncommitted_changes: true' in your ~/.evergreen.yml file to include uncommitted changes. Continue? (Y/n) y problem getting git metadata: can't get git version: can't parse git version number from version string 'git version 2.15.0.windows.1' E:\workspace\mongo>git --version git version 2.15.0.windows.1 ",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12747","08/10/2020 14:52:19",2,"Handle empty running task on host page ""On the host page, there may not be a running host. When there is no running task, it should display Current Task: _none._ !image-2020-08-10-15-28-58-974.png!""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12748","08/10/2020 14:54:44",2,"Add data and tests for host events ""Add data and tests that test a host that has events attached to both a host id and a tag. When we create an intent host it has an ID like “evg-abcdefgh”, and then when we create the actual host we now have an ID like “i-1234"""". The host intent ID is saved as Tag` in the actual host. The current query should cover both, but we need to add a test that specifically includes events that happened to the host intent. ""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12755","08/10/2020 16:18:46",2,"Task Queue: Route ""* /task-queue * Just going to /task-queue page will redirect user to default distro queue * e.g. /task-queue => /task-queue/macos-1012 * Selected task represented in route * /task-queue/:distroId/:taskId Regardless of distro id * ‘Task Queue` page title always present * ‘Search for Distro’ input always present""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12758","08/10/2020 16:22:37",2,"Tasks Queue: Task table filtering ""* Can filter by task ID * Can filter by revision""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12762","08/10/2020 19:44:30",1,"Clean up interns' virtual workstations ""EVG-8183 provided an API for MANA to clean up unexpirable resources (e.g. virtual workstations) left behind when someone leaves the company. Since MANA doesn't call it yet we need to manually clean up the hosts/volumes left behind by the summer interns.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12763","08/10/2020 21:00:51",1,"Commit queue command and patch show truncated commit message ""Using evergreen version 2020-08-06, I ran The commit message is truncated from """"SERVER-50221 Use shared_ptr to manage NoLimitSorter and InMemIterator in-memory data"""" to """"SERVER-50221 Use shared_ptr to manage NoLimitSorter and InMemIterator"""", and it appears this way in the Evergreen UI as well. It appears that the point of truncation is the point where the commit message rolls over to the next line on [GitHub|https://github.com/mongodb/mongo/commit/6d307c337d5d76e5471e0676fc255cab15aef61a].""","evergreen commit-queue merge --project mongodb-mongo-master src/mongo/db/sorter/sorter.cpp | 41 ++++++++++++++++++++++------------------- 1 file changed, 22 insertions(+), 19 deletions(-) 190e83f93f SERVER-50221 Use shared_ptr to manage NoLimitSorter and InMemIterator in-memory data This is a summary of the patch to be submitted. Continue? (y/n): y Patch successfully created. ID : 5f31a602a4cf474b2600ee5d Created : 2020-08-10 19:54:42.903 +0000 UTC Description : Commit Queue Merge: 'SERVER-50221 Use shared_ptr to manage NoLimitSorter and InMemIterator' into 'mongodb/mongo:master' Build : https://evergreen.mongodb.com/patch/5f31a602a4cf474b2600ee5d?redirect_spruce_users=true Status : created Queue position is 0.",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12764","08/11/2020 04:24:18",0,"Client should return non-zero status when user aborts patch submission at the uncommitted changes prompt ""I typically run {code} evergreen patch ... && git branch server-xxxxx-evgN {code} or similar so that I can keep track of my patch builds. This works nicely if I change my mind and answer """"n"""" to the """"This is a summary of the patch"""" prompt, or if the patch submission fails for some reason. However, if I answer """"n"""" to the """"Uncommitted changes are omitted from patches by default"""" prompt, then the exit code is still 0. I would expect it to be non-zero, since the patch submission didn't run to successful completion. The success exit status means the git branch/tag (or whatever followup action I wanted) is still performed, even though the patch wasn't submitted. This is counter-intuitive given the way that the other prompt works. Usually when I say """"n"""" to this prompt it's because I just want to run """"git status"""" to see what the uncommitted changes are. """," evergreen patch ... && git branch server-xxxxx-evgN $ evergreen version 2020-08-06 $ evergreen patch -d 'foo' Uncommitted changes are omitted from patches by default. Use the '--uncommitted, -u' flag or set 'patch_uncommitted_changes: true' in your ~/.evergreen.yml file to include uncommitted changes. Continue? (Y/n) n $ echo $? 0 $ evergreen patch -d 'foo' Uncommitted changes are omitted from patches by default. Use the '--uncommitted, -u' flag or set 'patch_uncommitted_changes: true' in your ~/.evergreen.yml file to include uncommitted changes. Continue? (Y/n) y src/mongo/db/repl/SConscript | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) 49f95061530 SERVER-50002 separate unittest binary for ReplicaSetAwareService This is a summary of the patch to be submitted. Continue? (y/n): n patch aborted $ echo $? 1 ",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12765","08/11/2020 15:39:47",2,"Exclude dev releases from the download feed ""Dev releases should be available in the repo but should not be advertised publicly for downloading since it's meant for internal use.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12766","08/11/2020 15:51:48",1,"Return 200 error on list endpoints with no resources attached ""List endpoints like /resource_A/a_id/B_resources should not 404 when there are no B resources attached to an existing resource_A. They should return a 200 with an empty list. requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://evergreen.mongodb.com/rest/v2/tasks/mms_code_health_E2E_NDS_DRY_RUN_c93893c3366dc473a6eaacade37416d56ef25b16_20_08_11_13_22_13/tests""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12767","08/11/2020 16:46:30",1,"Do not remove temporary file prematurely when updating evergreen binary ""This will cause updates via {{evergreen get-update --install}} to fail.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12769","08/11/2020 20:43:42",2,"Poplar should enable env vars for sensitive information ""Some variables that are included in the upload report, such as the bucket api key, can only be added via the json report file. This file is often exported with the rest of the DSI artifacts which probably shouldn't have such sensitive information. We should enable env vars for these elements.""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12777","08/12/2020 18:58:44",2,"Pushed tag not triggering new version ""After configuring the {{mongo-tools}} project for tag-triggered versions and pushing a tag called {{test-tag}}, I saw that the tag was added to the [existing version|https://evergreen.mongodb.com/version/mongo_tools_d3aec0abfc02b849946ab154e1cda1884c0d06d5] corresponding to the tagged commit, but no new version was created. For reference, see the attached screenshot of the tag-triggered versions config. More info from a slack conversation with [~annie.black]: {quote}it looks like we weren't able to find any builds, but i see that release is a valid build so i'm not sure why it wasn't found{quote}""","",0,0,0,0,0,0,1,0,0,1,0,0,0,0,0
+"EVG-12779","08/12/2020 21:35:18",2,"Stop using git's -c option in git.get_project ""git version < 2.1.2 doesn't support it""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12784","08/13/2020 16:22:25",2,"Restricted project mms-auto-code-change evergreen project appears in dropdown for Cory Mintz ""My understanding is that a restricted project should only be viewable by the admins of the project. That must be incorrect. How can I change who can view this project? If I cannot change it myself, I need removed everyone except the admins from view access.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12788","08/13/2020 22:34:02",2,"Re-execution of system failure did not activate blocked tasks ""I have a task which claims to be blocked. However, it depends only on a compile task, which successfully completed. The compile task appears to have encountered system failures, and took four tries to successfully complete. Is this expected for dependent tasks to remain blocked? https://evergreen.mongodb.com/task/mongodb_mongo_master_debian92_aggregation_patch_424f10ec4bc7b333bd682c7490fc6147688fd0ec_5f3558913066156e86024d1b_20_08_13_15_15_09##comparehashes=424f10ec4bc7b333bd682c7490fc6147688fd0ec&threads=all""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12789","08/14/2020 19:23:59",2,"gantt chart for tasks incorporating wait times ""Characterize system-introduced lag for SLAs. Potential Metrics: slowdown: runtime on a loaded system divided by runtime on a dedicated system by task tree or by build wait time: Time from scheduling to dispatch problem: different schedulers function differently on this""","",0,0,1,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12790","08/14/2020 19:54:43",1,"Can't start stopped hosts from the old UI ""Clicking the start button returns a 400""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12798","08/17/2020 17:32:02",1,"Add new metadata fields to ArtifactVersion ""{{ArtifactVersion}}'s {{ProductionRelease}} and {{DevelopmentRelease}} are only for the old versioning scheme. The download feed will have {{ContinuousRelease}} and {{LTSRelease}} for the new versioning scheme. Also, we should fix the JSON tags for {{Current}}. It doesn't seem link any project is currently using these fields so far.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12799","08/17/2020 18:48:03",1,"Fix semantics of IsDevelopmentBuild for NewMongoDBVersion ""It should be false for any new version that is not """"pre"""" or ends in """"-"""" or has """"~"""".""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12801","08/17/2020 21:29:49",2,"Handle undefined hosts more gracefully ""!image-2020-08-17-16-28-42-844.png!""","",0,0,0,1,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12804","08/18/2020 16:00:22",2,"validate_commit_message task in the commit queue does not work with a commit message across two lines ""I had difficulty with [a recent commit|https://github.com/mongodb/mongo/commit/456b0e3ff9c319f494f6defcf62315cd17f8d3a6] where I originally had the commit message across two lines so that each line was <=100 chars, like so: SERVER-49731 Create an Ident base class for RecordStore and SortedDataInterface and thread a shared_ptr through to the KVDropPendingIdentReaper on drop collection/index. When I tried to use evergreen commit-queue merge --project mongodb-mongo-master to send the patch to the commit queue, the message got completely garbled: Commit Queue Merge: '=?UTF-8?q?SERVER-49731=20Create=20an=20Ident=20base=20class=20for?=' into 'mongodb/mongo:master' (https://spruce.mongodb.com/version/5f3bd794e3c331627f992d36/tasks) I changed it to a single really long line, and it finally worked fine: https://spruce.mongodb.com/version/5f3bd94de3c331627f993096/tasks Is this expected behavior? Or a bug? What format commit message should I be using? This is the first commit message where this has been a problem. I've committed patches with two-line messages before that the commit queue turned into one line, which was not what I wanted but did get through the commit queue.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12807","08/18/2020 19:32:53",1,"Add region credentials to distros in prod ""This is dependent on the build script being deployed to prod.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12809","08/18/2020 21:04:02",1,"Set up CI testing for bond ""Currently, testing is all manual. We should also set up golangci-lint as the preferred linter instead of gometalinter.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12810","08/18/2020 21:38:52",2,"Ensure feeds have expected output for first continuous release ""Run some tests and manually check to ensure that the full.json and current.json will look as they're expected in preparation for the actual release. The other feeds that will be generated should also be spot-checked to ensure they'll produce the expected output. Since the versions are just pulled from git tags in the mongo repo, it seems fine to just add the first continuous version to that list and see if it produces the expected feed.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12814","08/19/2020 15:37:52",2,"Investigate Failed to acquire IX lock transaction errors ""This is causing commits to be missed in rare cases""","",0,0,1,0,0,0,0,0,0,1,0,0,1,0,0
+"EVG-12824","08/20/2020 15:15:53",1,"Clean up makefiles ""1. Remove things used for goEnv that were necessary solely for gometalinter (mostly the PATH muddling). 2. Should also experiment with removing goEnv entirely and just setting environment variables using Make's {{export}} feature. 3. Clean up build directory creation.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12825","08/20/2020 15:16:18",1,"Store regular patches in diff format ""[This commit|https://github.com/evergreen-ci/evergreen/commit/4321ae58c139be58a2370a4c0d6e62f9236ec3b2] causes patches to be stored in patch format. This breaks tests relying on git.get_project not committing the changes. We should hold off on storing patches this way until the tests can handle the changes being committed.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12826","08/20/2020 15:17:59",1,"Store regular patches in patch format ""Once tests have been adjusted to handle committed changes we should go back to storing patches in patch (mbox) format.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12827","08/20/2020 15:59:17",2,"Mbox BodyReader cannot handle lines longer than 4096 bytes ""msg.BodyReader().Read calls nextLine(), which calls ReadSlice('\n'), which relies on the default bufio buffer size of 4096 bytes. This means that lines longer than 4096 bytes cause patches to fail.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12829","08/20/2020 16:28:35",2,"Spawnhost setup scripts cd to non-existent directory on RHEL6.2 ""The user for RHEL is root and the root user's home directory is at /root. When we [prepend|https://github.com/evergreen-ci/evergreen/blob/b043069916c03fbdedc843526ee204eec8e38481/units/util.go#L52] the script with {{cd /home/}}, it fails because the directory /home/root doesn't exist.""","",0,0,1,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12838","08/21/2020 18:16:28",1,"Merge new versioning scheme logic ""Once the mongo-release-tools logic is code-complete and ready for the first continuous release, it should be merged into the master branch.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12840","08/21/2020 19:54:33",1,"Handle dev releases for feeds and ensure continuous releases get release notes ""Ensure that: - Development releases appear in full.json but not in current.json. - Continuous releases get their own release notes, but Docs will have to update us on what the release notes URL will be for them.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12843","08/24/2020 14:29:58",2,"Failures: coverage on Coverage [Evergreen @ a5450e38] (TestBackgroundSuite, TestMaxHeartbeats) "" h2. [coverage failed on Coverage|https://evergreen.mongodb.com/task/evergreen_coverage_coverage_a5450e38434901d38c62b11f5243f3c2bc88c91e_20_08_21_20_35_21/0] Host: [ec2-34-235-152-164.compute-1.amazonaws.com|https://evergreen.mongodb.com/host/i-049d55f089ef5f68d] Project: [Evergreen|https://evergreen.mongodb.com/waterfall/evergreen] Commit: [diff|https://github.com/evergreen-ci/evergreen/commit/a5450e38434901d38c62b11f5243f3c2bc88c91e]: EVG-12481 Don't panic (#3918) | 21 Aug 20 20:35 UTC Evergreen Subscription: ; Evergreen Event: *TestBackgroundSuite* - [Logs|https://evergreen.mongodb.com/test_log/5f4032a82fbabe0a1f64ef69] | [History|https://evergreen.mongodb.com/task_history/evergreen/coverage?revision=a5450e38434901d38c62b11f5243f3c2bc88c91e#/TestBackgroundSuite=fail] *TestMaxHeartbeats* - [Logs|https://evergreen.mongodb.com/test_log/5f4032a82fbabe0a1f64ef69] | [History|https://evergreen.mongodb.com/task_history/evergreen/coverage?revision=a5450e38434901d38c62b11f5243f3c2bc88c91e#/TestMaxHeartbeats=fail] [Task Logs | https://evergreen.mongodb.com/task_log_raw/evergreen_coverage_coverage_a5450e38434901d38c62b11f5243f3c2bc88c91e_20_08_21_20_35_21/0?type=T] ""","",0,0,0,0,1,0,0,0,0,1,0,0,0,0,0
+"EVG-12845","08/24/2020 15:28:08",1,"Provide API for a patch's diff ""Currently If a task wants to get its patch it needs to get it itself from git diff. We should provide an endpoint for the task to get it from. This will reduce reliance on how Evergreen applies the patch.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12851","08/24/2020 18:30:21",1,"commit queue panics ""https://mongodb.slack.com/archives/C866SR2LR/p1598289724003000""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12852","08/24/2020 21:50:32",2,"improve offboarding users by modifying expiration time ""Manual testing has shown that offboarding users takes more time than our route allows, given the one minute time out. To make this as successful as possible, I believe this should be moved to a job. Potentially if terminating an unexpirable instance fails, we should log critical, otherwise the unexpirable volume may never be terminated. Title updated to reflect the comments.""","",0,0,0,1,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12873","08/25/2020 21:57:51",1,"Use latest version of golangci-lint ""The last go1.9 version of golangci-lint is buggy. It's probably better to lint with the latest linter version since go1.9 compatibility doesn't matter much.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12874","08/25/2020 22:55:35",1,"Add link to task page that goes to /task-queue ""link only exists if task is on the queue""","",0,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12877","08/25/2020 23:42:06",1,"Ensure that if parameter YAML defaults are changed for PRs, these are used ""This includes PRs.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12876","08/25/2020 23:42:06",2,"Allow default parameters to be stored in the project YAML ""These should overwrite project variables. Should allow for a default value and a description.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12881","08/25/2020 23:42:07",2,"Allow user to pass in parameters from the CLI, using either KEY=VALUE pairs or a file ""These parameters should also be stored with the patch and exposed via API.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12878","08/25/2020 23:42:07",2,"Allow users to maintain project-specific default parameters in their local client configuration. ""This will just add onto the existing Project struct in local configs. This should overwrite defaults in the project YAML.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12883","08/25/2020 23:42:08",1,"Ensure parameters are documented where relevant ""And potentially parameters should have their own section, to explain the hierarchy.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12882","08/25/2020 23:42:08",1,"Add --parameters option to the evergreen list command ""This will display the description and defaults for the parameters defined in the YAML for a project.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12888","08/26/2020 18:57:41",2,"host.create should log the distros or AMI it creates ""There have been several cases where it has been difficult to debug a problem with host.create, because it required reading a complex project config to figure out what distro was spawned. host.create should log the distro or AMI it is spawning to the agent logs.""","",0,1,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12896","08/27/2020 15:25:21",1,"Use proper gRPC error codes ""Jasper currently just uses the {{github.com/pkg/errors}} to return errors to the gRPC service. However, because of this, it will always return an internal RPC error for any of these cases. We should use gRPC's error formatting {{google.golang.org/grpc/status}} since it supports returning specific gRPC statuses.""","",0,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"EVG-12899","08/27/2020 19:19:38",1,"Change LastCommunication Format on /host/ ""Change the format from Last Communication: Jul 20, 2020 9:08:41 am to Last Communication: a month ago""","",0,0,1,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12903","08/28/2020 15:45:59",2,"Way to exclude tasks from git tag triggered versions ""Similar to how we have """"patchable: false"""" it would be great to have a similar flag to disallow things from running on git tag triggered versions instead of just messing with regex ""","",1,0,0,0,0,0,0,0,0,1,0,0,0,0,0
+"EVG-12907","08/28/2020 17:55:58",2,"Add the add note button and functionality to the build baron UI "" Main Ticket [Design|https://mongodb.invisionapp.com/share/C2WKQWRWG5Y#/screens/415337715] ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12910","08/28/2020 17:56:53",2,"Display related Jira tickets in the build baron ""Fetch related Jira Tickets data on the ui. [Main Ticket|https://jira.mongodb.org/browse/EVG-12459] [Design|https://mongodb.invisionapp.com/share/C2WKQWRWG5Y#/screens/415337715] ""","",1,0,0,0,0,0,0,0,0,0,0,0,0,1,0
+"EVG-12913","08/31/2020 22:23:51",2,"Implement required scripting environment methods for CLI ""They have not been implemented so far.""","",0,0,0,1,0,0,0,0,0,0,0,0,1,0,0