Hi,
This morning I deployed a different version of _dumbwin32proc.py to all
of our Win7 (talos-r3-w7-*) and WinXP machines (talos-r3-xp-*).
This different version allows the machines to kill processes that
currently could not be killed. For instance, when we asked a job to be
interrupted. It would simply not happen.
I have been watching the tree to see if I can catch any regression for
those 2 pools of machines. If you see anything out of the ordinary
please let me know in bug 798170 [1].
So far, this change (if it works as expected) should improve our wait
times. For instance, jobs that failed to kill a process and instead it
times out OR when a job would get canceled but the job would still run
all the way. Let's hope that this works as expected and see better wait
times for those 2 pools.
Next week I will do the same for the Win64 builders and the Win8 machines.
Best regards,
Armen [2]
[1] https://bugzilla.mozilla.org/show_bug.cgi?id=798170
[2] https://mozillians.org/en-US/u/armenzg
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform