On Fri, 14 Feb 2025 at 02:00, Kevin Fenzi <kevin@xxxxxxxxx> wrote:
On Thu, Feb 13, 2025 at 12:15:16PM -0500, Dusty Mabe wrote:
>
>
> On 2/13/25 11:42 AM, Kevin Fenzi wrote:
> > I agree with downthread folks that that seems like way too high a
> > failure rate to enable gating on. However, a few questions if I can:
> >
> > Is this reporting to bodhi for all these components?
> > Off hand checking a few I don't see any results from this?
>
> Hmm. Anytime a test gets run it should report results to bodhi. One thing
> that might be tripping us up is these metrics are based on the activity in
> our matrix channel https://matrix.to/#/#jenkins-coreos:fedoraproject.org
> and IIUC we might have an issue where the reported RPM in the matrix message
> that gets sent out *may* not be the actual RPM we're tracking (i.e. if an
> update has 5 packages in it the reported failure may be against an RPM we're
> actually not tracking, but the overall test and reported failure to bodhi
> are valid). This is probably how emacs got in the list of failures.
Ah. Makes sense.
> > Wouldn't a good first step be to enable (non gating) these to show up
> > there so maintainers/you can be more aware and help reduce problems?
>
> Is there an example where you think the test should have reported results
> to bodhi, but didn't? It's probably what I mentioned above.
I looked again and I missed that these were just mixed in with the rest.
Somehow I was expecting another section. I assume they are the:
coreos.cosa.build-and-test and the like tests?
But if you look at say:
https://bodhi.fedoraproject.org/updates/FEDORA-2025-9f28dbb79a
I don't actually see any coreos* results there?
ndbkit is not a package where we trigger the cosa tests again, see the full list https://github.com/coreos/coreos-ci/blob/main/bodhi-testing.yaml
It is I believe a similar to what Dusty described earlier, where an update contains more than one package. I could make the script that I use to collect the data to be smarter to look only at the packages we are triggering tests on.
> > Also, I see in the table a few packages have really high failure rate.
> > (nbdkit, makedumpfile, etc). Perhaps fixing these would lower the entire
> > failure rate a good deal?
>
> Indeed, we're improving all the time!
:)
kevin
--
_______________________________________________
devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx
Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
-- _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
