Project

General

Profile

Actions
Copy link

Bug #61890

open

osd/numa_affinity: osd use cpuset is inconsistent with cgroup alloc

Added by yite gu 11 months ago. Updated 10 months ago.

Status:
Fix Under Review
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(RADOS):
Pull request ID:
52307
Crash signature (v1):
Crash signature (v2):

Description

cgroup allow cpuset to osd pod as blow:

# cat /sys/fs/cgroup/cpuset/kubepods.slice/kubepods-besteffort.slice/kubepods-besteffort-poda3e235fd_f1a0_49cf_b372_43e95708d117.slice/cri-containerd-9e375c52aac64897c5225ab8fb1bda0568b87f55b43815f7bdf22e74fa1259bc.scope/cpuset.cpus
3,35-63,100-127

but osd process really use cpuset as blow:
# taskset -c -p 948575
pid 948575's current affinity list: 3

Actions
Copy link
 #2

Updated by Radoslaw Zarzynski 10 months ago

  • Status changed from New to Fix Under Review
  • Pull request ID set to 52307
Actions
Copy link

Also available in: Atom PDF