-
Bug
-
Resolution: Done
-
Major
-
9.3.1.Final, 9.4.0.Final
The protocol servers never pass the configured worker threads to the netty event loops which default to 2 * available cpus (as counted by the JDK, which might not take cgroup/cfs quotas into account)..
- causes
-
CLOUD-2489 [JDG7] CPU cores are not reflecting set quotas while running in a pod
- Closed