[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
Re: [jakartaee-tck-dev] Should we update some of the TCK jobs to send email on failure?
|
On Fri, Feb 7, 2020 at 6:36 PM Bill Shannon <bill.shannon@xxxxxxxxxx> wrote:
>
> Scott Marlow wrote on 2/7/20 6:16 AM:
>
>
>
> On Fri, Feb 7, 2020, 8:35 AM Scott Marlow <smarlow@xxxxxxxxxx> wrote:
>>
>> On Fri, Feb 7, 2020 at 8:16 AM Tom Jenkinson <tom.jenkinson@xxxxxxxxxx> wrote:
>> >
>> > Are these different to the emails we get reporting failures?
>>
>> It should be a fairly generic Jenkins email message, I haven't seen
>> one yet, as TCK failures have occurred yet.
>
>
> No further errors have occurred yet.
>
> I'm really, really, really hoping that's because we fixed the gratuitous failures that were causing many of the CI jobs to fail randomly!
I'm not sure if I abused the system but I tried running two concurrent
TCK jobs [1][2] and they did fail. Normally, the
jakartaeetck-nightly-build-master job prevents this from happening, as
we only allow one concurrent run, however, since [2] is a separate
job, both [1][2] did run concurrently.
I'm not super worried about being able to run both [1][2] at the same
time, but am still wondering if our configuration could be improved to
allow running both [1][2] at the same time, where we ensure that
neither job overuses the available resources.
I'm not sure as to how we have cpu [3] + memory [4] resources are
configured for our TCK jobs, which could be related to failures that I
saw when trying to test the 8.0.2 TCK release. I'm also not sure how
many (virtual) CPUs + memory per CPU, we have for running each TCK
test.
I will check to see if we need to update other TCK jobs to prevent
concurrent runs, so at least us users won't as easily hit this
stability issues.
So, in summary, I think that the current limitation is that we can
only run one full Platform TCK run concurrently at a time, which we
should be careful about if we do another EE 8.0 TCK release, while
also doing EE 9 TCK testing.
Scott
[1] https://ci.eclipse.org/jakartaee-tck/job/jakartaeetck-nightly-build-master/53/
[2] https://ci.eclipse.org/jakartaee-tck/job/jakartaeetck-nightly-build-802/8/
[3] https://kubernetes.io/docs/tasks/configure-pod-container/assign-cpu-resource/
[4] https://kubernetes.io/docs/tasks/configure-pod-container/assign-memory-resource/
