-
Notifications
You must be signed in to change notification settings - Fork 404
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
OperationProcessor queue may be left closed while the container is in RUNNING state #6021
Comments
I think that we have now a clearer picture of this problem. The sequence of events that leads to this problem is as follows:
Given that for the |
This problem has been detected despite the previous PR, so reopening. |
Describe the bug
Under a high load scenario, we sporadically face an issue in which:
A Bookie restarts/disconnects due to high load, which lead the Bookkeeper client to throw
BKNotEnoughBookiesException
and then some Segment Containers restart.After that, a Segment Store seems that have all the containers recovered:
Could it be a problem in the initialization/shutdown sequence that leaves sometimes the Container as RUNNING but
OperationProcessor
queue as closed?To Reproduce
Requires a high load scenario in which Bookies are close to saturation. In this case, SLTS is also enabled. If one or more Bookies are restarted or disconnected, it may be possible to see this issue.
Screenshots
n/a
Additional information
The problem is not persistent: once we manually restart the Segment Store impacted, everything comes back to normal.
The text was updated successfully, but these errors were encountered: