Project

General

Profile

Actions

Bug #66124

closed

mds: choose a new batch head in request_cleanup() instead

Added by Xiubo Li almost 2 years ago. Updated 5 months ago.

Status:
Resolved
Priority:
Normal
Assignee:
Category:
Correctness/Safety
Target version:
% Done:

100%

Source:
Q/A
Backport:
squid,reef,quincy
Regression:
No
Severity:
2 - major
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(FS):
MDS
Labels (FS):
Pull request ID:
Tags (freeform):
backport_processed
Fixed In:
v19.3.0-2510-gc8f71d6495
Released In:
v20.2.0~2780
Upkeep Timestamp:
2025-11-01T01:26:49+00:00

Description

Try to improve this and choose a new batch head in request_cleanup() instead.

More detail please see tracker: https://tracker.ceph.com/issues/65536 and ceph PR comments: https://github.com/ceph/ceph/pull/56941#discussion_r1602421868.


Related issues 5 (0 open5 closed)

Copied from CephFS - Bug #65536: mds: after the unresponsive client was evicted the blocked slow requests were not successfully cleaned upResolvedXiubo Li

Actions
Copied to CephFS - Backport #66363: squid: mds: choose a new batch head in request_cleanup() insteadResolvedXiubo LiActions
Copied to CephFS - Backport #66563: squid: mds: choose a new batch head in request_cleanup() insteadDuplicateXiubo LiActions
Copied to CephFS - Backport #67162: reef: mds: choose a new batch head in request_cleanup() insteadResolvedXiubo LiActions
Copied to CephFS - Backport #67163: quincy: mds: choose a new batch head in request_cleanup() insteadResolvedXiubo LiActions
Actions #1

Updated by Xiubo Li almost 2 years ago

  • Copied from Bug #65536: mds: after the unresponsive client was evicted the blocked slow requests were not successfully cleaned up added
Actions #2

Updated by Xiubo Li almost 2 years ago

  • Status changed from In Progress to Fix Under Review
  • Pull request ID set to 57553
Actions #3

Updated by Leonid Usov almost 2 years ago

  • Component(FS) deleted (quiesce)
Actions #4

Updated by Patrick Donnelly almost 2 years ago

  • Status changed from Fix Under Review to Pending Backport
Actions #6

Updated by Xiubo Li almost 2 years ago

  • Copied to Backport #66363: squid: mds: choose a new batch head in request_cleanup() instead added
Actions #7

Updated by Upkeep Bot almost 2 years ago

  • Copied to Backport #66563: squid: mds: choose a new batch head in request_cleanup() instead added
Actions #8

Updated by Xiubo Li over 1 year ago

  • Backport changed from squid to squid,reef,quincy
Actions #9

Updated by Xiubo Li over 1 year ago

Backporting to quincy and reef could resolve the crash issue in https://tracker.ceph.com/issues/66967#note-8

Actions #10

Updated by Upkeep Bot over 1 year ago

  • Copied to Backport #67162: reef: mds: choose a new batch head in request_cleanup() instead added
Actions #11

Updated by Upkeep Bot over 1 year ago

  • Copied to Backport #67163: quincy: mds: choose a new batch head in request_cleanup() instead added
Actions #12

Updated by Upkeep Bot over 1 year ago

  • Tags (freeform) set to backport_processed
Actions #13

Updated by Konstantin Shalygin over 1 year ago

  • Status changed from Pending Backport to Resolved
  • % Done changed from 0 to 100
Actions #14

Updated by Upkeep Bot 9 months ago

  • Merge Commit set to c8f71d64951c8f60953caab3f431162c1f898c17
  • Fixed In set to v19.3.0-2510-gc8f71d64951
  • Upkeep Timestamp set to 2025-06-27T03:18:34+00:00
Actions #15

Updated by Upkeep Bot 8 months ago

  • Fixed In changed from v19.3.0-2510-gc8f71d64951 to v19.3.0-2510-gc8f71d6495
  • Upkeep Timestamp changed from 2025-06-27T03:18:34+00:00 to 2025-07-14T16:44:57+00:00
Actions #16

Updated by Upkeep Bot 5 months ago

  • Released In set to v20.2.0~2780
  • Upkeep Timestamp changed from 2025-07-14T16:44:57+00:00 to 2025-11-01T01:26:49+00:00
Actions

Also available in: Atom PDF