Issues causing end-2-end test failures on https://openshift.io - Since Dec 2017

[maxandersen]

Note added January 3 - The issues listed below are affecting, but not 100% blocking the execution of the EE tests. The tests are running and are seeing a success rate of ~80-90%:

• https://ci.centos.org/view/Devtools/job/devtools-test-end-to-end-ts-openshift.io-runTest/
• https://ci.centos.org/view/Devtools/job/devtools-test-end-to-end-ts-openshift.io-chequickstartTest/

Tracking issues for known issues blocking end-2-end to run on production (https://openshift.io):

Jenkins/Forge/Pipeline related issues

• [x] Jenkins pipeline not created reliably #1194
• [x] Che fails to create workspaces #1304
• [x] Forge fails to create quickstarts #1503
• [x] EBS related issues seen - Che/Jenkins pods fail to start - intermittent #1520
• [x] Intermittent build failures, no route to host for local nexus #1536
• [x] Intermittent build failures, timeout - socket closed #1537
• [x] On prod-preview, creating a new app/quickstart - the app wizard hangs #1550
• [x] Promote to stage/run failing on prod-preview #1725
• [ ] Seeing slower than usual times to start a build pipeline #1731
• [x] Intermittent maven error results in build pipeline failures #1735
• [x] Jenkins pod not starting #1855
• [ ] Intermittent slow thruput on prod-preview #1904
• [ ] Intermittent long delays (>1 hour) starting build pipeline #1926
• [x] On production, "promote" button fails, blocks deploy to run #2034
• [x] On prod-preview, creating new quickstart fails #2035
• [x] On production, pipeline page corrupted #2276
• [x] On production - Jenkins idler fails - build pipeline cannot run #2341
• [ ] Random failures of OK button in creating new quickstarts #2369
• [ ] Maven throttling stops builds #2408
• [ ] No route to Jenkins host #2423
• [ ] Maven socket closed and pipeline builds fail #2426
• [ ] Jenkins slave never connects #2427

Che related issues

• [x] Intermittent error creating Che workspace - Che stuck in 'starting workspace' state #1573
• [x] Intermittent error creating Che workspace - Unable to initialize IDE #1585
• [x] Intermittent error creating Che workspace - No workspace running #1615
• [x] Intermittent error creating Che workspace - Workspace agent is no longer responding #1616
• [ ] Creating Che workspace could fail after logout / workspace idling (KC token used by Che is invalidated) #1733
• [x] Che is not able to create workspaces on https://openshift.io - #1860
• [x] Creating a Che workspace fails on production - #1705
• [x] Incorrect stack used in Che for quicklstarts, "run" command missing #1873
• [x] Unable to create a new Che workspace on prod-preview #1917
• [x] Cannot create new Che workspace on production #2308
• [ ] Random/recurring failure creating workspace #2420

Issues related to tests running on each PR - all blocking bugs are resolved - only remaining issue is the time (~20 min) required to perform a build of Fabric8-ui

• [x] PR to trigger test execution https://jenkins.cd.test.fabric8.io/job/fabric8-ui/job/fabric8-ui/view/change-requests/job/PR-2341/
• [x] Fabric8-ui build is not fully functioning due to CORS issue - opened at https://gitlab.cee.redhat.com/dtsd/housekeeping/issues/909
• [x] Executing the tests failed due to a login issue - UI changed, workaround at https://github.com/fabric8io/fabric8-test/pull/338
• [x] Test code fails for status icon - workaround https://github.com/fabric8io/fabric8-test/pull/339

[maxandersen]

@ldimaggi please help keep this task list above uptodate on what blocks end-2-end and tick them off when you are unblocked.

[ibuziuk]

e2e tests are failing on both prod & prod-preview since Jan 24th due to https://github.com/openshiftio/openshift.io/issues/2034. Shouldn't this issue be marked as SEV1 ?

[ldimaggi]

Sev1 is for production only, not prod-preview,

This is a bad bug, but, the user is not 100% blocked as they can promote from either the Jenkins build log or from OSO.

But - yes - we should fix this ASAP...

[ppitonak]

@ldimaggi can we close this issue?

[ldimaggi]

This issue lists some very old issues - if we want to have a parent JIRA for all new failures, I'd recommend opening a new issue. Thx!