* [TEST] Google's Netdev-CI: missing results
@ 2026-10-23 15:52 Matthieu Baerts
2026-11-20 9:08 ` Matthieu Baerts
0 siblings, 1 reply; 3+ messages in thread
From: Matthieu Baerts @ 2026-10-23 15:52 UTC (permalink / raw)
To: Sheena Mohan; +Cc: Netdev, Willem de Bruijn, Jakub Kicinski
Hi Sheena,
I'm working on the Netdev CI, and I notice that the runners on Google
side are frequently missing results:
https://netdev.bots.linux.dev/flakes.html?tn-needle=google.*vlan-py&min-flip=0&br-pfx=net-next-hw
(Look at the non-coloured blocks.)
Is it normal? Do you mind looking at that please? :)
Note that when an active runner doesn't process a branch, we cannot mark
the whole batch being validated as "not causing regression" because it
is lacking some validations. So when this happens, the end result is
marked as "pending". That's not an issue when it is once in a while, but
here, it looks "too frequent" not to be reported.
Cheers,
Matt
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [TEST] Google's Netdev-CI: missing results
2026-10-23 15:52 [TEST] Google's Netdev-CI: missing results Matthieu Baerts
@ 2026-11-20 9:08 ` Matthieu Baerts
2026-11-20 22:19 ` Pin-yen Lin
0 siblings, 1 reply; 3+ messages in thread
From: Matthieu Baerts @ 2026-11-20 9:08 UTC (permalink / raw)
To: Sheena Mohan; +Cc: Netdev, Willem de Bruijn, Jakub Kicinski, Pin-yen Lin
Hi Sheena,
(+ cc Pin-yen)
On 23/10/2026 17:52, Matthieu Baerts wrote:
> I'm working on the Netdev CI, and I notice that the runners on Google
> side are frequently missing results:
>
> https://netdev.bots.linux.dev/flakes.html?tn-needle=google.*vlan-py&min-flip=0&br-pfx=net-next-hw
>
> (Look at the non-coloured blocks.)
>
> Is it normal? Do you mind looking at that please? :)
>
> Note that when an active runner doesn't process a branch, we cannot mark
> the whole batch being validated as "not causing regression" because it
> is lacking some validations. So when this happens, the end result is
> marked as "pending". That's not an issue when it is once in a while, but
> here, it looks "too frequent" not to be reported.
Lately, it looks like the situation got worst: 3 days without results.
Because it is blocking all the other HW tests, I temporarily disabled
the two Google runners: results (if any) are still fetched, but ignored.
Do you mind improving the situation, please?
Cheers,
Matt
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [TEST] Google's Netdev-CI: missing results
2026-11-20 9:08 ` Matthieu Baerts
@ 2026-11-20 22:19 ` Pin-yen Lin
0 siblings, 0 replies; 3+ messages in thread
From: Pin-yen Lin @ 2026-11-20 22:19 UTC (permalink / raw)
To: Matthieu Baerts; +Cc: Sheena Mohan, Netdev, Willem de Bruijn, Jakub Kicinski
Hi Matthieu,
On Fri, Nov 20, 2026 at 1:08 AM Matthieu Baerts <matttbe@kernel.org> wrote:
>
> Hi Sheena,
>
> (+ cc Pin-yen)
>
> On 23/10/2026 17:52, Matthieu Baerts wrote:
> > I'm working on the Netdev CI, and I notice that the runners on Google
> > side are frequently missing results:
> >
> > https://netdev.bots.linux.dev/flakes.html?tn-needle=google.*vlan-py&min-flip=0&br-pfx=net-next-hw
> >
> > (Look at the non-coloured blocks.)
> >
> > Is it normal? Do you mind looking at that please? :)
> >
> > Note that when an active runner doesn't process a branch, we cannot mark
> > the whole batch being validated as "not causing regression" because it
> > is lacking some validations. So when this happens, the end result is
> > marked as "pending". That's not an issue when it is once in a while, but
> > here, it looks "too frequent" not to be reported.
>
> Lately, it looks like the situation got worst: 3 days without results.
> Because it is blocking all the other HW tests, I temporarily disabled
> the two Google runners: results (if any) are still fetched, but ignored.
>
> Do you mind improving the situation, please?
Thanks for looping me in.
This was caused by two conflicting patches ([1], [2]) on the mailing
list. [2] removed .resizing from struct htable, which [1] references.
Our runners currently build from a full distro kernel config (which
enables CONFIG_IP_SET), so kernel compilation failed and no test
results were uploaded.
Results for google-gve are being reported again today now that [2] is
no longer in the testing branches. Sorry for the inconvenience and for
blocking the other HW tests.
We will update our testing infrastructure to use a minimal kernel to
avoid unrelated build breakage in the future.
[1]: https://lore.kernel.org/all/20261117092731.14656-1-onleaf322@gmail.com/
[2]: https://lore.kernel.org/all/20261117101018.523826-1-fuchen.dust@gmail.com/
Thanks and regards,
Pin-yen
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-11-20 22:19 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-10-23 15:52 [TEST] Google's Netdev-CI: missing results Matthieu Baerts
2026-11-20 9:08 ` Matthieu Baerts
2026-11-20 22:19 ` Pin-yen Lin
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.