Details
-
Bug
-
Resolution: Duplicate
-
Major
-
Lustre 2.4.2
-
None
-
Lustre 2.4.2-14chaos (see github.com/chaos/lustre)
ZFS OSD
-
3
-
15756
Description
We had an MDS crash with the following assertion:
ldlm_lock.c:770:ldlm_lock_addref()) ASSERTION( lock != ((void *)0) )
The previous few minutes contained many lock timeouts and some node reconnections.
The backtrace for the mdt_02_049 that hit the assertion was:
ldlm_lock_addref mdt_reint_open mdt_reconstruct_open mdt_reconstruct mdt_reint_internal mdt_intent_reint mdt_intent_policy ldlm_lock_enqueue ldlm_handle_enqueue0 mdt_enqueue mdt_handle_common mds_regular_handle ptlrpc_server_handle_request ptlrpc_main
We were running lustre version 2.4.2-14chaos (see github.com/chaos/lustre).
We cannot provide logs or crash dumps from this system.
Attachments
Issue Links
Activity
Fix Version/s | New: Lustre 2.5.0 [ 10295 ] | |
Resolution | New: Duplicate [ 3 ] | |
Status | Original: Open [ 1 ] | New: Resolved [ 5 ] |
Affects Version/s | New: Lustre 2.4.2 [ 10605 ] |
Assignee | Original: WC Triage [ wc-triage ] | New: Niu Yawei [ niu ] |