merlimat commented on PR #22841:
URL: https://github.com/apache/pulsar/pull/22841#issuecomment-2148700163
> It might not give a guarantee for all scenarios
That's equal to no guarantees
> I don't see any harm to keep it multi threads for metadata-store and if
you don't want to keep in all usecases then I can make it configurable from
service configuration.
It does bring multiple problems:
1. Masks problems that must be eradicated at the source. The problem is
still there, and it will still be happening, just it becomes harder to track.
2. Multiple threads will bring in ordering issue across callbacks
3. Potentially more mutex contention
> we are wasting lot of time to keep our system stable with such issues. and
this fix helps us to keep it stable in such regressions where extra thread
helps to prevent race-condition and because of that avoid the deadlock.
I understand, though the problem was introduced by masking the blocking call
in the interface you introduced a while back.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
