[LU-16781] account jobids to original process Created: 27/Apr/23  Updated: 27/Apr/23

Status: Open
Project: Lustre
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Improvement Priority: Minor
Reporter: Andreas Dilger Assignee: WC Triage
Resolution: Unresolved Votes: 0
Labels: None

Issue Links:
Related
is related to LU-16765 Allow longer JobID names Open
is related to LU-16766 Combine some kernel process names for... Resolved
Rank (Obsolete): 9223372036854775807

 Description   

For user-initiated threads like ll_sa_* these should be associated with the UID/GID/PID of the original task for %u, %g, and %p, so that they can be accounted correctly on the servers.

For jobid_name_is_valid() we may consider to reduce the exclusions for kernel processes. Firstly, this exclusion doesn't always working properly, since kworker and ll_sa tasks still show up in the server stats. Secondly, this hides the real presence of RPCs sent to the server, so job_stats are not showing the full picture of what is generating the load.

The client needs to save the originating comm/UID/GID/PID for the work that is done (e.g. cache in inode when it is read/written). This should already be happening, but doesn't seem to be used consistently.


Generated at Sat Feb 10 03:29:57 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.