All,
I feel I owe an explanation for why GitHub checks were failing so
frequently this week.
I noticed checks were waiting for a while so I reimaged braggi09 and
adami03 from CentOS8 to Ubuntu Focal. The goal was to get make check
and API tests pushed through more quickly.
That led to me find out that not all build deps were getting installed
automatically (https://tracker.ceph.com/issues/55059)
On top of that, GitHub had intermittent outages a couple times this
week. This means that sometimes GitHub wasn't triggering builds in
Jenkins AND if the builds *did* get triggered, sometimes Jenkins wasn't
able to report that build result back to GitHub.
PLUS all the docs builds started failing due to
https://github.com/sphinx-doc/sphinx/issues/10291
The docs checks are temporarily not required as I merge all the PRs that
have been approved but just failing that check.
The sphinx issue has PR for master and Laura is working on backporting
other docs fixes to quincy. It'll need to go back to Pacific too I believe.
master: https://github.com/ceph/ceph/pull/45644
quincy: https://github.com/ceph/ceph/pull/45627
I'm sure this isn't all encompassing and there were some other minor
issues (e.g., https://github.com/ceph/ceph/pull/45647) but I think we're
getting close to stable again.
Thanks for your patience,
--
David Galloway
Principal Systems Administrator
Ceph Engineering
_______________________________________________
Dev mailing list -- dev@xxxxxxx
To unsubscribe send an email to dev-leave@xxxxxxx