CI congestion/starvation

[Ben Cooksley]

On Sat, Feb 14, 2026 at 5:47 AM Volker Krause <vkrause at kde.org> wrote:

> On Freitag, 13. Februar 2026 16:42:03 Mitteleuropäische Normalzeit Sune
> Vuorela wrote:
> > On 2026-02-13, Harald Sitter <sitter at kde.org> wrote:
> > > On Fri, Feb 13, 2026 at 12:37 PM Ben Cooksley <bcooksley at kde.org>
> wrote:
> > >> Resource utilisation wise, i've not looked into whether there has
> been a
> > >> significant bump in the number of jobs, but over the past year some
> > >> additional CD support has been added so that indicates some trouble
> > >> there.>
> > > Going off on a tangent: if the resources aren't sufficient for the
> > > development of our flagship products (ruqola is not one, nor is
> > > messagelib, but also I don't know what either do without looking them
> > > up so maybe they are crucial to something 🤷‍♂️) then we need to put
> > > more resources up.
> >
> > ruqola is a chat client for rocket.chat whatever that is. It seems that
> > a linux build takes 10 minutes so I'm a bit surprised that it shows up.
> >
> > Messagelib is quite in the heart of our email offerings. A linux build
> > seems to take 4-5 minutes, so I'm also a bit surprised that that shows
> > up here.
>
> Right, messagelib is essentially KMail. Build time is only part of the
> problem
> there, it's quite test-heavy as all the email crypto stuff is in there.
>

The per job stats for messagelib:

         name          |    time_used    | job_count
-----------------------+-----------------+-----------
windows_qt610         | 24:52:33.347276 |        73
suse_tumbleweed_qt610 | 18:28:10.854189 |        91
suse_tumbleweed_qt611 | 18:13:52.270202 |        91
freebsd150_qt610      | 18:08:18.798915 |        73
cppcheck              | 02:39:29.341629 |        76
pre-commit            | 01:40:32.695585 |        54
json-validation       | 00:20:55.701217 |        83
clang_format          | 00:18:01.056173 |        56
xml                   | 00:05:28.430049 |        27
yaml                  | 00:05:25.533611 |        27

Having a quick look through seems like messagelib has some flaky tests,
leading to developers hitting retry unnecessarily on jobs, particularly for
suse_qt611?

Windows also has a couple of tests that timeout, leading to those jobs
taking longer than they should by about 6 minutes per run, which adds up
fast when you run 73 times (about 7 of the 25 hours above are wasted
waiting for a timeout).
Of the 85 hours spent by messagelib on CI - 8% of them were on three tests
on Windows that timed out.


>
> Regards,
> Volker


Thanks,
Ben
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-devel/attachments/20260214/fc591c5c/attachment.htm>