You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The test_prospector.py:Test.test_harvester_limit test is flaky. The fail can be reproduced by running it in a while loop like this:
while nosetests test_prospector.py:Test.test_harvester_limit; do echo .; done
On my computer, this fails after 10 or 20 iterations.
Looking at the logs and code, it seems to be a race condition between checking p.harvesters.Len() and p.harvesters.Start(). We could serialise the whole startHarvester() method with a new lock, but I'm slightly worried about dead locks, because the harvester registry has it's own lock. So I'm opening a ticket for now to separate this from #5449.
The text was updated successfully, but these errors were encountered:
This fixes a race where the size of registry's internal map was
used to count the number of harvesters, but this map is handled
in a different goroutine.
Closeselastic#5458
The
test_prospector.py:Test.test_harvester_limit
test is flaky. The fail can be reproduced by running it in a while loop like this:On my computer, this fails after 10 or 20 iterations.
Looking at the logs and code, it seems to be a race condition between checking
p.harvesters.Len()
andp.harvesters.Start()
. We could serialise the wholestartHarvester()
method with a new lock, but I'm slightly worried about dead locks, because the harvester registry has it's own lock. So I'm opening a ticket for now to separate this from #5449.The text was updated successfully, but these errors were encountered: