Bug#1088254: tracker.debian.org: seems to be severely lagging behind ci.debian.net for (at least) some packages

Hi Paul,

On Thu, Nov 28, 2024 at 12:44:15PM +0100, Paul Gevers wrote:
> On 11/28/24 11:00, Julian Gilbey wrote:
> > > We had issues on ci.d.n last week, which also led to britney1 bailing out
> > > because there were no results.
> > 
> > I guess it was those problems that caused the issue, then.  (Just to
> > clarify, in this case, the results hadn't been transferred from ci.d.n
> > to britney1/britney2 after 3 days, and showed no signs of ever
> > transferring.)
> 
> I think the issue only lasted half a day, one day at most. So 3 days is
> unexplained.
> 
> For avoidance of misconception, you are talking about tests scheduled by the
> britney user and using the current trigger version?

I'll avoid mentioning "britney user" because I'm not exactly sure what
that entails; I'll instead describe what I did and observed from my
end.

(1) I uploaded jupyter-notebook 6.4.13-4 to unstable on 21 Nov 2024,
and received an email "Accepted ... into unstable"; this is visible
on tracker and is timestamped Thu, 21 Nov 2024 20:09:16 +0000

(2) I monitored tracker.debian.org for the next few days, and noticed
that all the tests were saying "Test in progress" for a while.

(3) I reported this on 25 Nov 2024 when I saw that tracker was still
saying "Test in progress" for all tests on all affected packages
(emperor, jupyter-notebook and sagenb-export) but discovered that
ci.debian.net listed all (or perhaps almost all) of the tests as
having passed or being neutral on 22 Nov.

I hope this makes sense!

Best wishes,

   Julian