-
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
Infrastructure Flake on Jenkins when building openshift/node image: [Errno -1] repomd.xml does not match metalink for epel #8571
Infrastructure Flake on Jenkins when building openshift/node image: [Errno -1] repomd.xml does not match metalink for epel #8571
Comments
This has gotten worse ...
@tdawson did you say this was just EPEL mirrors being out of sync with the source? Can we mitigate failures like this in CI? |
Mirrors are certainly possible. @tdawson Do you have any particular recommendations or advice on the best approach? |
@stevekuznetsov Can you try adding a |
It has nothing to do with the yum caches - due to the way the repo is managed updates will always be racy. It would be a lot less racy if 2 copies of the RPM data were cached and the RPMs were sync'd before the repodata. |
I'd say use openshifts mirrors, they have epel mirrored and are only updated once a day, so you won't get a weird update halfway through your build. Instead of installing epel-release, do the following RUN wget -O /etc/yum.repos.d/epel7.repo https://mirror.openshift.com/mirror/epel/epel7.repo |
We can address this by switching to the This will be blocked on @smarterclayton implementing that secrets feature. |
You do not need the certs/key for that area https://mirror.openshift.com/mirror/epel/ Since there is nothing private in that repo, you do not need a certificate to access it. If you are getting failures when you are accessing them, please let me know. |
@tdawson ah, incredible. I assumed I would need them due to some other repos on the domain needing them. @smarterclayton since our internal mirror doesn't need the secrets, can you see anything wrong with simply disabling the upstream EPEL repo in the RHEL base image that I think we use for all of our builds? |
Slightly different, but still this:
|
When building the openshift/node docker image:
https://ci.openshift.redhat.com/jenkins/job/test_pull_requests_origin_integration/208/console
The text was updated successfully, but these errors were encountered: