Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Backport smoke/conformance test harness changes to release-1.1 from master #18686

Conversation

karlkfi
Copy link
Contributor

@karlkfi karlkfi commented Dec 14, 2015

The following changes (hopefully) make it feasible to run release-1.1 branch conformance tests on master code in mesos environments. This should also fix mesos smoke tests on cherry-pick PRs to release-1.1.

(This only affects smoke/conformance tests. So e2e tests are not required.)

Karl Isenberg added 2 commits December 14, 2015 14:37
- Skip [Skipped] tests when testing for conformance kubernetes#17180
- Add cluster/test-conformance.sh script kubernetes#17997
- Support passing args to cluster/test-smoke.sh kubernetes#15404
- Support annotated tags in conformance test version detection kubernetes#17839
@k8s-github-robot
Copy link

Labelling this PR as size/L

@k8s-github-robot k8s-github-robot added the size/L Denotes a PR that changes 100-499 lines, ignoring generated files. label Dec 14, 2015
@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 14, 2015

Mesosphere CI

pull/18686/head:

master with CONFORMANCE_BRANCH=pull/18686/head:

@k8s-bot
Copy link

k8s-bot commented Dec 14, 2015

GCE e2e build/test failed for commit ffa5655.

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

Conformance Results:

pull/18686/head:

[16:34:07][Step 7/7] Summarizing 3 Failures:
[16:34:07][Step 7/7] 
[16:34:07][Step 7/7] [Fail] SchedulerPredicates [It] validates resource limits of pods that are allowed to run [Conformance] 
[16:34:07][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/_output/local/go/src/k8s.io/kubernetes/test/e2e/scheduler_predicates.go:87
[16:34:07][Step 7/7] 
[16:34:07][Step 7/7] [Fail] Kubectl client Guestbook application [It] should create and stop a working application [Conformance] 
[16:34:07][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/_output/local/go/src/k8s.io/kubernetes/test/e2e/kubectl.go:1010
[16:34:07][Step 7/7] 
[16:34:07][Step 7/7] [Fail] Networking [It] should function for intra-pod communication [Conformance] 
[16:34:07][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/_output/local/go/src/k8s.io/kubernetes/test/e2e/networking.go:180
[16:34:07][Step 7/7] 
[16:34:07][Step 7/7] Ran 72 of 175 Specs in 3704.877 seconds
[16:34:07][Step 7/7] FAIL! -- 69 Passed | 3 Failed | 2 Pending | 101 Skipped >>> testContext.KubeConfig: /root/.kube/config
[16:34:07][Step 7/7] Skipping SSH core dump, which is not implemented for --- FAIL: TestE2E (3704.91s)
[16:34:07][Step 7/7] FAIL
[16:34:07][Step 7/7] 
[16:34:07][Step 7/7] Ginkgo ran 1 suite in 1h1m45.322904307s
[16:34:07][Step 7/7] Test Suite Failed

master with CONFORMANCE_BRANCH=pull/18686/head:

[16:32:17][Step 7/7] Summarizing 3 Failures:
[16:32:17][Step 7/7] 
[16:32:17][Step 7/7] [Fail] SchedulerPredicates [It] validates resource limits of pods that are allowed to run [Conformance] 
[16:32:17][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/conformance/_output/local/go/src/k8s.io/kubernetes/test/e2e/scheduler_predicates.go:141
[16:32:17][Step 7/7] 
[16:32:17][Step 7/7] [Fail] EmptyDir volumes [It] should support (non-root,0777,default) [Conformance] 
[16:32:17][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/conformance/_output/local/go/src/k8s.io/kubernetes/test/e2e/util.go:1173
[16:32:17][Step 7/7] 
[16:32:17][Step 7/7] [Fail] Networking [It] should function for intra-pod communication [Conformance] 
[16:32:17][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/conformance/_output/local/go/src/k8s.io/kubernetes/test/e2e/networking.go:159
[16:32:17][Step 7/7] 
[16:32:17][Step 7/7] Ran 72 of 175 Specs in 2624.442 seconds
[16:32:17][Step 7/7] FAIL! -- 69 Passed | 3 Failed | 2 Pending | 101 Skipped >>> testContext.KubeConfig: /root/.kube/config
[16:32:17][Step 7/7] Skipping SSH core dump, which is not implemented for --- FAIL: TestE2E (2624.47s)
[16:32:17][Step 7/7] FAIL
[16:32:17][Step 7/7] 
[16:32:17][Step 7/7] Ginkgo ran 1 suite in 43m44.875094434s
[16:32:17][Step 7/7] Test Suite Failed

SchedulerPredicates test is fixed by #18348 (pending merge).

The other failures need inspection, but are probably related to code or test changes, not the test harness.

@karlkfi karlkfi changed the title WIP Backport smoke/conformance test harness changes to release-1.1 Backport smoke/conformance test harness changes to release-1.1 Dec 15, 2015
@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

@k8s-bot e2e test this

@karlkfi karlkfi removed their assignment Dec 15, 2015
@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

@kubernetes/goog-testing Can one of you guys look this over and give it a thumb?

It's a manual backport because #17511 was already half backported and cherry-picking 6.5 PRs sounded like a PITA. No tests were changed, just harness scripts.

Full e2e shouldn't be affected/required, but I kicked it off for completeness.

@karlkfi karlkfi changed the title Backport smoke/conformance test harness changes to release-1.1 Backport smoke/conformance test harness changes to release-1.1 from master Dec 15, 2015
@k8s-bot
Copy link

k8s-bot commented Dec 15, 2015

GCE e2e build/test failed for commit ffa5655.

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

Ran master with CONFORMANCE_BRANCH=pull/18686/head again. Looks like 2/3 of the errors were flakes. This time, only the expected SchedulerPredicates test failed (since then, #18348 has been merged).

[18:45:49][Step 7/7] Summarizing 1 Failure:
[18:45:49][Step 7/7] 
[18:45:49][Step 7/7] [Fail] SchedulerPredicates [It] validates resource limits of pods that are allowed to run [Conformance] 
[18:45:49][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/_output/local/go/src/k8s.io/kubernetes/test/e2e/scheduler_predicates.go:139
[18:45:49][Step 7/7] 
[18:45:49][Step 7/7] Ran 79 of 206 Specs in 2792.012 seconds
[18:45:49][Step 7/7] FAIL! -- 78 Passed | 1 Failed | 1 Pending | 126 Skipped Dec 15 02:45:49.040: INFO: >>> testContext.KubeConfig: /root/.kube/config
[18:45:49][Step 7/7] 
[18:45:49][Step 7/7] Skipping SSH core dump, which is not implemented for --- FAIL: TestE2E (2802.05s)
[18:45:49][Step 7/7] FAIL
[18:45:49][Step 7/7] 
[18:45:49][Step 7/7] Ginkgo ran 1 suite in 46m42.552753932s
[18:45:49][Step 7/7] Test Suite Failed

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

A rerun of the conformance tests on this PR shows 2/3 failures were flakes, and 2 more flakes popped up. SchedulerPredicates is the only consistently failing test. We'll likely cherry-pick #18348

[19:22:02][Step 7/7] Summarizing 3 Failures:
[19:22:02][Step 7/7] 
[19:22:02][Step 7/7] [Fail] SchedulerPredicates [It] validates resource limits of pods that are allowed to run [Conformance] 
[19:22:02][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/_output/local/go/src/k8s.io/kubernetes/test/e2e/scheduler_predicates.go:87
[19:22:02][Step 7/7] 
[19:22:02][Step 7/7] [Fail] Examples e2e [Example]ClusterDns [It] should create pod that uses dns [Conformance] 
[19:22:02][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/_output/local/go/src/k8s.io/kubernetes/test/e2e/examples.go:505
[19:22:02][Step 7/7] 
[19:22:02][Step 7/7] [Fail] Docker Containers [It] should be able to override the image's default commmand (docker entrypoint) [Conformance] 
[19:22:02][Step 7/7] /go/src/github.com/GoogleCloudPlatform/kubernetes/_output/local/go/src/k8s.io/kubernetes/test/e2e/util.go:1173
[19:22:02][Step 7/7] 
[19:22:02][Step 7/7] Ran 72 of 175 Specs in 4427.685 seconds
[19:22:02][Step 7/7] FAIL! -- 69 Passed | 3 Failed | 2 Pending | 101 Skipped >>> testContext.KubeConfig: /root/.kube/config
[19:22:02][Step 7/7] Skipping SSH core dump, which is not implemented for --- FAIL: TestE2E (4427.72s)
[19:22:02][Step 7/7] FAIL
[19:22:02][Step 7/7] 
[19:22:02][Step 7/7] Ginkgo ran 1 suite in 1h13m48.114844558s
[19:22:02][Step 7/7] Test Suite Failed

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

From Google's Jenkins, here's the e2e failure. Either it's a flake or release-1.1 doesn't pass conformance. Will re-run to check.

Summarizing 1 Failure:

[Fail] Services [It] should serve multiport endpoints from pods [Conformance] 
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/test/e2e/service.go:1048

Ran 110 of 175 Specs in 1047.911 seconds
FAIL! -- 109 Passed | 1 Failed | 2 Pending | 63 Skipped 

Ginkgo ran 1 suite in 17m36.672718671s
Test Suite Failed

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

@k8s-bot e2e test this

@k8s-bot
Copy link

k8s-bot commented Dec 15, 2015

GCE e2e build/test failed for commit ffa5655.

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

Jenkins GCE seems to be hosed.

ERROR: (gcloud.compute.firewall-rules.create) Some requests did not succeed:
 - Quota 'FIREWALLS' exceeded.  Limit: 100.0

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

Ran master with CONFORMANCE_BRANCH=pull/18686/head again, now that #18348 is merged.

100% conformance!

[11:11:21][Step 7/7] Ran 79 of 206 Specs in 2906.263 seconds
[11:11:21][Step 7/7] SUCCESS! -- 79 Passed | 0 Failed | 1 Pending | 126 Skipped Dec 15 19:11:21.932: INFO: >>> testContext.KubeConfig: /root/.kube/config
[11:11:21][Step 7/7] 
[11:11:21][Step 7/7] Skipping SSH core dump, which is not implemented for PASS
[11:11:21][Step 7/7] 
[11:11:21][Step 7/7] Ginkgo ran 1 suite in 48m36.832110792s
[11:11:21][Step 7/7] Test Suite Passed

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

@k8s-bot e2e test this

@k8s-bot
Copy link

k8s-bot commented Dec 15, 2015

GCE e2e build/test failed for commit ffa5655.

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 15, 2015

Same problem. It appears google's quota is filled, and I can't get ahold of an on-call to fix it.

@kubernetes/goog-testing You guys have any control over this?

ERROR: (gcloud.compute.firewall-rules.create) Some requests did not succeed:
 - Quota 'FIREWALLS' exceeded.  Limit: 100.0

@karlkfi
Copy link
Contributor Author

karlkfi commented Dec 16, 2015

@k8s-bot e2e test this

@k8s-bot
Copy link

k8s-bot commented Dec 16, 2015

GCE e2e test build/test passed for commit ffa5655.

@lavalamp lavalamp assigned ixdy and unassigned lavalamp Jan 4, 2016
@lavalamp
Copy link
Member

lavalamp commented Jan 4, 2016

Sorry for delay, I've been out of the office. @ixdy, can you take a look or find an owner?

@ikehz ikehz assigned ikehz and unassigned ixdy Jan 4, 2016
@ikehz
Copy link
Contributor

ikehz commented Jan 4, 2016

I'll have a look.

@ikehz ikehz added the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Jan 4, 2016
@ikehz
Copy link
Contributor

ikehz commented Jan 4, 2016

LGTM.

@k8s-github-robot
Copy link

@k8s-bot test this

Tests are more than 48 hours old. Re-running tests.

@k8s-bot
Copy link

k8s-bot commented Jan 4, 2016

GCE e2e test build/test passed for commit ffa5655.

@k8s-github-robot
Copy link

@k8s-bot test this [submit-queue is verifying that this PR is safe to merge]

@k8s-bot
Copy link

k8s-bot commented Jan 5, 2016

GCE e2e test build/test passed for commit ffa5655.

@k8s-github-robot
Copy link

Automatic merge from submit-queue

k8s-github-robot pushed a commit that referenced this pull request Jan 5, 2016
@k8s-github-robot k8s-github-robot merged commit b9dfea3 into kubernetes:release-1.1 Jan 5, 2016
@karlkfi karlkfi deleted the release-1.1-mesos-ci-backport branch January 5, 2016 00:25
shyamjvs pushed a commit to shyamjvs/kubernetes that referenced this pull request Dec 1, 2016
…-ci-backport

Auto commit by PR queue bot
shouhong pushed a commit to shouhong/kubernetes that referenced this pull request Feb 14, 2017
…-ci-backport

Auto commit by PR queue bot
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
area/test area/test-infra lgtm "Looks good to me", indicates that a PR is ready to be merged. size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

8 participants