Actions
Bug #615
closedosd: improve op+journal throttling
% Done:
0%
Source:
Tags:
Backport:
Regression:
Severity:
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
Currently we block first, then take locks, then update the throttle accounting. This makes things racy, because a bunch of threads could pile up, wait, and then all wake up once a small amount is available and get submitted. The result is bursty. But fixing it properly will be a bit tricky...
Actions