-
Notifications
You must be signed in to change notification settings - Fork 69
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
Fix flaky TestMultipleClientsWithMixedLabelsAndExpectFailure
#599
Fix flaky TestMultipleClientsWithMixedLabelsAndExpectFailure
#599
Conversation
/find-flaky-tests Run: https://github.com/grafana/dskit/actions/runs/11237399709 |
3b58d5f
to
f76ff6b
Compare
fail-fast: false | ||
strategy: | ||
fail-fast: false |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'll get it right sometime... I managed to start a run anyways since "workflow dispatch" seems to use the workflow config from the branch, while a workflow triggered from issue comments takes the workflow config from main (while checking out the branch to run tests)
This failed in a bunch of runs in #598 When it fails, it has 3 updates, but they're all from the same member. I believe what we want to check is that the cluster never gets joined by members with labels
f76ff6b
to
7464c1e
Compare
#599 went in a wrong direction, so I revert it here Instead, to fix the `TestMultipleClientsWithMixedLabelsAndExpectFailure` test, I add a couple more labeled members, that way, even if we reach 3 updates (which happened some times), we'll never get to 5
#599 went in a wrong direction, so I revert it here: it makes all `TestMultipleClients` tests flaky since not all members are necessarily joined after the updates Instead, to fix the `TestMultipleClientsWithMixedLabelsAndExpectFailure` test, I add a couple more labeled members, that way, even if we reach 3 updates (which happened some times), we'll never get to 5 with a single member Also, try to fix the `TestTLSServerWithLocalhostCertWithClientCertificateEnforcementUsingClientCA1` test by closing GRPC connections. Maybe this (golang/go#36700) is the issue?
#599 went in a wrong direction, so I revert it here: it makes all `TestMultipleClients` tests flaky since not all members are necessarily joined after the updates Instead, to fix the `TestMultipleClientsWithMixedLabelsAndExpectFailure` test, I add a couple more labeled members, that way, even if we reach 3 updates (which happened some times), we'll never get to 5 with a single member Also, try to fix the `TestTLSServerWithLocalhostCertWithClientCertificateEnforcementUsingClientCA1` test by closing GRPC connections. Maybe this (golang/go#36700) is the issue?
* Really fix flaky `TestMultipleClients` tests #599 went in a wrong direction, so I revert it here: it makes all `TestMultipleClients` tests flaky since not all members are necessarily joined after the updates Instead, to fix the `TestMultipleClientsWithMixedLabelsAndExpectFailure` test, I add a couple more labeled members, that way, even if we reach 3 updates (which happened some times), we'll never get to 5 with a single member Also, try to fix the `TestTLSServerWithLocalhostCertWithClientCertificateEnforcementUsingClientCA1` test by closing GRPC connections. Maybe this (golang/go#36700) is the issue? * Set `MaxIdleConnsPerHost` * New attempt to fix this test. Could it be aborted connections? * Retry RSTs
This failed in a bunch of runs in #598
When it fails, it has 3 updates, but they're all from the same member. I believe what we want to check is that the cluster never gets joined by members with labels