I can reliably reproduce this failure 5% of the time when the test is run by itself. So I am confused as to why we don't see periodic failures of this test in buildbot.
As a final test I ran test_timeout_long and test_timeout_short together to see if there was some interaction. Again I saw 1 failure out of 20.
After discussion with Gary I am classifying test_timeout_short as a spurious failure, marking it with an XXX referencing this bug, and disabling it for now.
I can reliably reproduce this failure 5% of the time when the test is run by itself. So I am confused as to why we don't see periodic failures of this test in buildbot.
As a final test I ran test_timeout_long and test_timeout_short together to see if there was some interaction. Again I saw 1 failure out of 20.
After discussion with Gary I am classifying test_timeout_short as a spurious failure, marking it with an XXX referencing this bug, and disabling it for now.