Project

General

Profile

Actions
Copy link

Bug #966

closed

mds: scatterstat error

Added by Sage Weil about 13 years ago. Updated over 7 years ago.

Status:
Can't reproduce
Priority:
High
Assignee:
Sage Weil
Category:
-
Target version:
-
% Done:

0%

Source:
Tags:
Backport:
Regression:
Severity:
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(FS):
Labels (FS):
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

- vstart d -n -x
 cfuse mnt
- cp /usr mnt/foo for a while (~15 mb of mds journal)
- stop the copy
- restart mds's


mds/CDir.cc: In function 'bool CDir::check_rstats()', in thread '0x7fe52f6e9700'
mds/CDir.cc: 225: FAILED assert(!g_conf.mds_debug_scatterstat || (get_num_head_items() == (fnode.fragstat.nfiles + fnode.fragstat.nsubdirs)))
 ceph version 0.25-534-g3f1e9b0 (commit:3f1e9b0f7b87f0d2113e72e33c328f647c3eb2ef)
 1: (ceph::__ceph_assert_fail(char const*, char const*, int, char const*)+0x53) [0xa5076a]
 2: (CDir::check_rstats()+0x758) [0x942d06]
 3: (CInode::finish_scatter_gather_update(int)+0x1718) [0x969b22]
 4: (Locker::scatter_writebehind(ScatterLock*)+0x273) [0x8fc077]
 5: (Locker::simple_sync(SimpleLock*, bool*)+0x3ee) [0x8fb340]
 6: (Locker::scatter_nudge(ScatterLock*, Context*, bool)+0x464) [0x8fcda4]
 7: (Locker::scatter_tick()+0x231) [0x8fd30b]
 8: (MDS::tick()+0x376) [0x76bae8]
 9: (MDS::C_MDS_Tick::finish(int)+0x32) [0x77f34e]
 10: (SafeTimer::timer_thread()+0x217) [0xa4dadd]
 11: (SafeTimerThread::entry()+0x1c) [0xa4e7b8]
 12: (Thread::_entry_func(void*)+0x23) [0x74ac29]
 13: (()+0x68ba) [0x7fe532e418ba]
 14: (clone()+0x6d) [0x7fe531ad602d]
 ceph version 0.25-534-g3f1e9b0 (commit:3f1e9b0f7b87f0d2113e72e33c328f647c3eb2ef)
 1: (ceph::__ceph_assert_fail(char const*, char const*, int, char const*)+0x53) [0xa5076a]
 2: (CDir::check_rstats()+0x758) [0x942d06]
 3: (CInode::finish_scatter_gather_update(int)+0x1718) [0x969b22]
 4: (Locker::scatter_writebehind(ScatterLock*)+0x273) [0x8fc077]
 5: (Locker::simple_sync(SimpleLock*, bool*)+0x3ee) [0x8fb340]
 6: (Locker::scatter_nudge(ScatterLock*, Context*, bool)+0x464) [0x8fcda4]
 7: (Locker::scatter_tick()+0x231) [0x8fd30b]
 8: (MDS::tick()+0x376) [0x76bae8]
 9: (MDS::C_MDS_Tick::finish(int)+0x32) [0x77f34e]
 10: (SafeTimer::timer_thread()+0x217) [0xa4dadd]
 11: (SafeTimerThread::entry()+0x1c) [0xa4e7b8]
 12: (Thread::_entry_func(void*)+0x23) [0x74ac29]
 13: (()+0x68ba) [0x7fe532e418ba]
 14: (clone()+0x6d) [0x7fe531ad602d]
*** Caught signal (Aborted) **

happens consistently. haven't tried with copying less data.

Actions
Copy link
 #1

Updated by Brian Chrisman about 13 years ago

Update this item when there's a fix submitted and I can retest my related issue fairly quickly.

Actions
Copy link
 #2

Updated by Greg Farnum about 13 years ago

  • Status changed from New to In Progress
  • Assignee set to Greg Farnum

I'll look into this today.

Actions
Copy link
 #3

Updated by Greg Farnum about 13 years ago

Ugh. My first attempt to reproduce this, the OSD crashed before the MDS did:

osd/ReplicatedPG.cc: 1211: FAILED assert(0 == "oi disagrees with stat")

Perhaps this OSD bug is the actual cause of the problem? (Which means that #989 is probably a different bug, actually...)

Actions
Copy link
 #4

Updated by Greg Farnum about 13 years ago

Oh, that assert actually occurred because of an ENOENT return code, not a stat mismatch. Pushed a fix, will try again.

Actions
Copy link
 #5

Updated by Greg Farnum about 13 years ago

  • Assignee changed from Greg Farnum to Sage Weil

I haven't been able to reproduce this even once. Assigning back to Sage per his request.

Actions
Copy link
 #6

Updated by Sage Weil about 13 years ago

  • Status changed from In Progress to Can't reproduce

Can't reproduce this one. Maybe it was resolved by commit:466306de3aeca22311993bf5a1955281499d751d.. or something earlier, given that greg couldn't hit it either!

Actions
Copy link
 #7

Updated by John Spray over 7 years ago

  • Project changed from Ceph to CephFS
  • Category deleted (1)
  • Target version deleted (v0.27)

Bulk updating project=ceph category=mds bugs so that I can remove the MDS category from the Ceph project to avoid confusion.

Actions
Copy link

Also available in: Atom PDF