-
Notifications
You must be signed in to change notification settings - Fork 4.7k
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
FAILURE: BUILD AN OPENSHIFT-ANSIBLE RELEASE #14122
Comments
I think I've fixed this by tagging stage 3.6.65.0 and master 3.6.66 |
Yeah, we need to be using that four-number workaround until we have a better solution, but this isn't necessarily a tooling or testing issue -- we just need to have a method for making unique tags. |
Well, we should use tools to handle the stage branching, so just making sure you guys are aware of that and we add .0 to the current version while branching and tagging stage. |
Is this problem solved? Are the people reporting this above doing so from old failures? Do we need to take action right now? |
Those are old failures. I fixed it at
|
seems like it's still there: https://ci.openshift.redhat.com/jenkins/job/test_pull_request_origin_extended_conformance_install_update/713/console |
Note that this happened exactly when the new tag was cut for OSE and the versions in error matches, so maybe just bad timing. |
#14322 hit it again - https://ci.openshift.redhat.com/jenkins/job/merge_pull_request_origin/853/ This seem to be serious :) |
On stage branch set rpm specfile version to 3.6.88.0, tag and pushed to github. We need to fix the stage branch tooling to append a fourth version segment when forking that branch. Looks like @jupierce built the jobs for that, assigning to him. |
@sdodson I think the only patch that will hold up through stagecut is adding a .0 to the openshift-ansible.spec in master. A change in the stage branch will be wiped out every time we build stage (since the o-a spec version is always replaced by the ose spec version). When stagecut ends and we return to master, the .0 will similarly be wiped out. |
Should be fixed with openshift-eng/aos-cd-jobs@ca7f5b2 |
Seen in #14853 https://ci.openshift.redhat.com/jenkins/job/merge_pull_request_origin/1111/
|
Fixed again. The present workaround relies on careful management of Jenkins job scheduling. Working on a more robust solution. |
@stevekuznetsov I think I asked this before during our discussion, but I'll lay it out there again. Addressing this completely outside of CI is tricky. I think a simple and effective approach would be for CI to sed append .1000 to any version it found in openshift-ansible.spec before doing tito tag. This is effectively what has to happen outside of CI to avoid this problem, but performing it outside of CI involves git commits and adds complexity to entering and exiting stagecut. |
Not the cleanest thing in the world but we can try that on Monday. |
We should now have a reliable fix in place:
|
Seen in https://ci.openshift.redhat.com/jenkins/job/test_pull_request_openshift_ansible_extended_conformance_install_with_status_check/260/consoleFull#-157296560358b6e51eb7608a5981914356
I think this happened because the tag is on the stage branch but this PR is against master?
The text was updated successfully, but these errors were encountered: