Closed tnozicka closed 7 years ago
@fabianofranz actually, this looks like it may be another instance of the --attach
issues we've seen in the past like https://github.com/openshift/origin/issues/12558 which was supposedly fixed in https://github.com/openshift/origin/pull/13669 but maybe just for test-end-to-end-docker.sh
.
@soltysh PTAL.
@pweil- correct, this is enabled only for test-end-to-docker.sh
, but that's the default we'll fallback to when docker is installed, which is what happens in our CI (look for ++ Docker is installed, running hack/test-end-to-end-docker.sh instead
). I'm worried this is back, but at this point in time I have no idea why, since the journald rate limiting is off (look for Turning off journald limits
). It's not a blocker for 3.6, but I'll be paying attention, since it looks like we're hitting it more and more often recently 😞
After IRC discussion with @pweil- I'll lower the prio. I need logs to verify what's going on and none of the current instances has those. It won't block the release in any way and I'll keep my eye open for it.
flaked on https://github.com/openshift/origin/pull/14745 too
@soltysh do you have everything you need in https://ci.openshift.redhat.com/jenkins/job/test_pull_request_origin_integration/4259/s3/ to debug this further?
also saw this here: https://github.com/openshift/origin/issues/15331 with a slightly different output.
and @mfojtik saw it here https://ci.openshift.redhat.com/jenkins/job/test_pull_request_origin/3264/ on a slightly different file (cli-with-token.log instead of cli-with-token2.log)
@bparees I can see:
I0719 15:31:37.212757 1 cached_discovery.go:88] failed to write cache to /root/.kube/172.30.0.1_443/authorization.openshift.io/v1/serverresources.json due to mkdir /root/.kube: permission denied
I0719 15:31:37.219134 1 cached_discovery.go:88] failed to write cache to /root/.kube/172.30.0.1_443/v1/serverresources.json due to mkdir /root/.kube: permission denied
@jupierce @kargakis @jhadvig anyone can check what are the permissions there?
@enj can you help investigate this? this seems like a number 1 flake atm.
I propose we disable this test for now to unblock merge/test queue: https://github.com/openshift/origin/pull/15363 objections?
no objections here, i just hit it again too.
There's a discussion about journald limits happening in https://github.com/openshift/origin/issues/14785
Is this just a dupe of the journald
issue?
Yeah, I'll close this one in favor of the other one.
github.com/openshift/origin/test/end-to-end/core.test/end-to-end/core.sh:389: executing 'cat '/data/src/github.com/openshift/origin/_output/scripts/test-end-to-end/logs/cli-with-token2.log'' expecting success and text 'system:serviceaccount:test:default' (from github.com_openshift_origin_test_end-to-end_core)
https://ci.openshift.redhat.com/jenkins/job/merge_pull_request_origin/1143/consoleFull
Stacktrace