Skip to content

Conversation

big-r81
Copy link
Contributor

@big-r81 big-r81 commented Aug 16, 2025

No description provided.

@big-r81 big-r81 requested a review from nickva August 21, 2025 15:45
@big-r81 big-r81 force-pushed the sm-v128-default-version branch 2 times, most recently from 9ac1000 to f73ea2d Compare August 28, 2025 07:19
@big-r81 big-r81 force-pushed the sm-v128-default-version branch from f73ea2d to 8d2508c Compare August 30, 2025 10:51
@nickva
Copy link
Contributor

nickva commented Aug 30, 2025

We'd have to make sure packaging works well with the change. Run a full ci build (git checkout to jenkins-* brach and then git commit --amend --no-edit && push).

I started one here: https://ci-couchdb.apache.org/job/jenkins-cm1/job/FullPlatformMatrix/job/jenkins-sm-128/1/pipeline-graph/

edit: looks good!

@big-r81
Copy link
Contributor Author

big-r81 commented Sep 1, 2025

Hey Nick,

I tried too, but I got an error for CentOS 8, because it's landed on the ARM64 worker and because it is using SM60, it failed. See https://ci-couchdb.apache.org/job/jenkins-cm1/job/FullPlatformMatrix/job/jenkins-sm-v128-default-version/1/pipeline-console/?selected-node=409

Another interesting case (but unrelated to the PR):
The same CI run is green, but one specific test fails: https://ci-couchdb.apache.org/job/jenkins-cm1/job/FullPlatformMatrix/job/jenkins-sm-v128-default-version/1/pipeline-console/

@nickva
Copy link
Contributor

nickva commented Sep 2, 2025

I tried too, but I got an error for CentOS 8, because it's landed on the ARM64 worker and because it is using SM60, it failed. See https://ci-couchdb.apache.org/job/jenkins-cm1/job/FullPlatformMatrix/job/jenkins-sm-v128-default-version/1/pipeline-console/?selected-node=409

Another interesting case (but unrelated to the PR): The same CI run is green, but one specific test fails: https://ci-couchdb.apache.org/job/jenkins-cm1/job/FullPlatformMatrix/job/jenkins-sm-v128-default-version/1/pipeline-console/

Hmm we may have to change the labeling scheme for how workers/hosts/guests find each other to run CI builds. Now we have a generic docker label because initially it meant they were all x86_64 basically, but we may have to find some other scheme: we may have to look the docker label but also at the os.arch (ex.aarch64) system property as well...

@big-r81
Copy link
Contributor Author

big-r81 commented Sep 2, 2025

Do we have an option for CentOS 8 to update SM to a version > 60?

@nickva
Copy link
Contributor

nickva commented Sep 2, 2025

Do we have an option for CentOS 8 to update SM to a version > 60?

Ideally we shouldn't have to. CentOS 8 should be shipping with its original SM major version and RH should be providing updates and security patches for if. If SM 60 comes from the EPEL repo then maybe there we can update it but if it's included in the base package set if we update it we'd be on the hook of maintaining our patches and security updates and such.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants