-
Notifications
You must be signed in to change notification settings - Fork 18.6k
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
Flaky test: TestAPISwarmNodeDrainPause #23516
Comments
test was temporarily disabled in #23526 |
----------------------------------------------------------------------
FAIL: docker_api_swarm_test.go:409: DockerSwarmSuite.TestApiSwarmNodeDrainPause
[d89110000] waiting for daemon to start
[d89110000] daemon started
docker_api_swarm_test.go:410:
d1 := s.AddDaemon(c, true, true)
check_test.go:221:
c.Assert(d.Init(aa, ""), check.IsNil)
... value *errors.errorString = &errors.errorString{s:"initializing swarm: invalid statuscode 406, \"{\\\"message\\\":\\\"this node is already part of a Swarm\\\"}\\n\""} ("initializing swarm: invalid statuscode 406, \"{\\\"message\\\":\\\"this node is already part of a Swarm\\\"}\\n\"")
---------------------------------------------------------------------- |
looks like this is supposed to be fixed via b38408f, but it's now failing consistently on power for the original reason. |
Also on s390x six DockerSwarm test cases fail:
Interestingly TestApiSwarmServicesCreateGlobal passes with gccgo and fails with golang. |
FYI: With commit 1bbac42 and go1.7 beta 2 on s390x:
2016-06-21-1bbac428de015a24a0724a9d0990d500c445dcad-test-k4.5-2cpu-go1.7b2.txt |
So the issue is that even though the daemon has an arch specific busybox cached, the swarm daemon automatically pulls down another one. While not intended to fix this issue, #23790 seems to solve the problem. @michael-holzheu can you see if that pr fixes the issue? alternative you can also temporarily change https://github.com/docker/docker/blob/master/integration-cli/docker_api_swarm_test.go#L665 to s390x/busybox:latest and that should do the trick as well. |
@tophj-ibm I can confirm that PR #23790 also fixes the issue on s390x. |
@thaJeztah Can we rename this issue since the test is no longer just flaky, but fails consistently? Or should we create another issue for the failing tests? |
@thaJeztah -- So, n/m. I misunderstood the conversation in PR #23790. And since that PR was merged (Thanks @tonistiigi and @vdemeester) and all 6 tests pass for power, z, and I'm assuming arm as well. 🎉 |
This happened in Z on #38500 too:
but looks to be quite rare error. |
Flaky test was renamed from Still looks like flakiness is an issue https://github.com/moby/moby/search?q=TestAPISwarmNodeDrainPause&type=Issues |
Looks to be caused by networking #38020 (comment)
Which is the same issue as #34051 (comment) (
Daemon logs attached:
|
Failing again on #39678 on PowerPC: https://ci.docker.com/public/job/moby/job/PR-39678/6/execution/node/208/log/
errors from daemon d1c7657d43a9c
errors from daemon dac8441f4d97e
|
This test was recently added, but appears to be flaky. I think @aluzzardi is already working on a fix in SwarmKit;
https://jenkins.dockerproject.org/job/Docker%20Master/label=ubuntu%20&&%20aufs/5741/console
The text was updated successfully, but these errors were encountered: