[LU-17183] sanity.sh test_411b: cgroups OOM on ARM Created: 11/Oct/23 Updated: 17/Nov/23 Resolved: 17/Nov/23 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Minor |
| Reporter: | Tim Day | Assignee: | Qian Yingjin |
| Resolution: | Duplicate | Votes: | 0 |
| Labels: | arm | ||
| Issue Links: |
|
||||||||||||
| Severity: | 3 | ||||||||||||
| Rank (Obsolete): | 9223372036854775807 | ||||||||||||
| Description |
|
The 411b test added in https://review.whamcloud.com/c/fs/lustre-release/+/50544/ regularly ran into OOM issues before the memory limit was increased significantly in https://review.whamcloud.com/c/fs/lustre-release/+/52610.
The memory required for this test to pass consistently on ARM is around 3x that of x86. This seems to suggest an issue (either with Lustre or the kernel) with cgroups on ARM. |
| Comments |
| Comment by Patrick Farrell [ 11/Oct/23 ] |
|
Thanks, Tim! |
| Comment by Xing Huang [ 28/Oct/23 ] |
|
2023-10-28: To be investigated. |
| Comment by Li Xi [ 17/Nov/23 ] |
|
After applied the patch from LU-17151, the problem is gone. |
| Comment by Qian Yingjin [ 17/Nov/23 ] |
|
After the patch in LU-17151 was merged into master branch with much larger memory limits on memcg, the failure on ARM does not happen any more.
|