On Mon, 28 Nov 2011 10:34:52 -0500 (EST)
Kamil Paral <kparal(a)redhat.com> wrote:
> I checked production and ... surprise surprise ... there are 4 jobs
> hanging from Nov 26! It seems that Nov 26 was a bad day.
>
>
http://autoqa.fedoraproject.org/results/238579-autotest/qa02.qa.fedorapro...
>
http://autoqa.fedoraproject.org/results/238665-autotest/qa07.qa.fedorapro...
>
http://autoqa.fedoraproject.org/results/238685-autotest/qa03.qa.fedorapro...
>
http://autoqa.fedoraproject.org/results/238799-autotest/qa01.qa.fedorapro...
>
> I'll abort them all. We will need to report bug in autotest. The
> reproducer will be tricky however.
Are we sure that this is an autotest bug? Something in the back of my
head says that we have the timeout before killing a job set to
something
crazy high.
Yes, I believe we did set it to something crazy high because the timer was buggy and some
jobs were being cancelled in a few minutes. I'll try to check.