Bugzilla

(In reply to Cosmin Sabou [:CosminS] from comment #70)
> I respectfully disagree. It was backed out two times but every time these failures were present also on your backed out pushes, two eg:
> first time it landed: http://tinyurl.com/yy7svmtj
> second time it landed: http://tinyurl.com/y5n679h6

Yes, and they are also present with any pushes prior in central.

> note how before your push those jobs were green (in some cases 100%) and then they started permafailling. this time they are intermittent but only just, for eg: http://tinyurl.com/y32u4zsk
And if you look at the try pushes I've done in the past. They were green too.
It's the nature of intermittents, they are intermittents :)

> also this failure wasn't even around the trees before your pushes as you can see from this 30 days failure rate: https://treeherder.mozilla.org/intermittent-failures.html#/bugdetails?startday=2019-04-27&endday=2019-05-27&tree=trunk&bug=1424433 so it looks like there are enough reasons to see how this is all connected to the modifications in bug 1550422. either this test is not important enough so it would be disabled or the bug you worked on would be backed out once again for that spike in intermittents.

That test fails locally without any of my changes with a 100% reproduction rate. It always times out.

As such I can't investigate a potential regression as there's no change of behaviour as far as I'm concerned.

I'm happy to disable the test, or better leave the author of the test look at it as I have no idea what this test is supposed to do nor can I assess what a normal behaviour is supposed to do seeing that this test always fail on my local win64 machines.

(In reply to Cosmin Sabou [:CosminS] from comment #70)
> I respectfully disagree. It was backed out two times but every time these failures were present also on your backed out pushes, two eg:
> first time it landed: http://tinyurl.com/yy7svmtj
> second time it landed: http://tinyurl.com/y5n679h6

Yes, and they are also present with any pushes prior in central.

> note how before your push those jobs were green (in some cases 100%) and then they started permafailling. this time they are intermittent but only just, for eg: http://tinyurl.com/y32u4zsk

And if you look at the try pushes I've done in the past. They were green too.
It's the nature of intermittents, they are intermittents :)

> also this failure wasn't even around the trees before your pushes as you can see from this 30 days failure rate: https://treeherder.mozilla.org/intermittent-failures.html#/bugdetails?startday=2019-04-27&endday=2019-05-27&tree=trunk&bug=1424433 so it looks like there are enough reasons to see how this is all connected to the modifications in bug 1550422. either this test is not important enough so it would be disabled or the bug you worked on would be backed out once again for that spike in intermittents.

That test fails locally without any of my changes with a 100% reproduction rate. It always times out.

As such I can't investigate a potential regression as there's no change of behaviour as far as I'm concerned.

I'm happy to disable the test, or better leave the author of the test look at it as I have no idea what this test is supposed to do nor can I assess what a normal behaviour is supposed to do seeing that this test always fail on my local win64 machines.