[LU-16692] replay-single: test_70c osp_fid_diff()) ASSERTION( fid_seq(fid1) == fid_seq(fid2) )

Related — Fri, 31 Mar 2023 11:46:34 +0000

This issue was created by maloo for eaujames <eaujames@ddn.com>

This issue relates to the following test suite run: https://testing.whamcloud.com/test_sets/cbcbb9b2-656c-44bd-b324-31c9dc39539e

Test session details:
clients: https://build.whamcloud.com/job/lustre-reviews/93413 - 4.18.0-348.7.1.el8_5.x86_64
servers: https://build.whamcloud.com/job/lustre-reviews/93413 - 4.18.0-348.23.1.el8_lustre.x86_64

The MDS crash with the following stack:

[11121.479852] LustreError: 567908:0:(osp_internal.h:538:osp_fid_diff()) ASSERTION( fid_seq(fid1) == fid_seq(fid2) ) failed: fid1:[0x280000bd1:0x2da4:0x0], fid2:[0x280000bd0:0x2da3:0x0]
[11121.482473] LustreError: 567908:0:(osp_internal.h:538:osp_fid_diff()) LBUG
[11121.483602] Pid: 567908, comm: tgt_recover_0 4.18.0-348.23.1.el8_lustre.x86_64 #1 SMP Thu Mar 2 00:54:25 UTC 2023
[11121.485233] Call Trace TBD:
[11121.485974] [<0>] libcfs_call_trace+0x6f/0x90 [libcfs]
[11121.486859] [<0>] lbug_with_loc+0x43/0x80 [libcfs]
[11121.487726] [<0>] osp_create+0x1dc/0xba0 [osp]
[11121.488585] [<0>] lod_sub_create+0x24a/0x4e0 [lod]
[11121.489407] [<0>] lod_striped_create+0x1a6/0x590 [lod]
[11121.490282] [<0>] lod_xattr_set+0x3f1/0x1300 [lod]
[11121.491160] [<0>] mdo_xattr_set+0x101/0x470 [mdd]
[11121.491967] [<0>] mdd_create_object+0x3cc/0x920 [mdd]
[11121.492821] [<0>] mdd_create+0xe62/0x1a30 [mdd]
[11121.493769] [<0>] mdt_reint_open+0x2d1a/0x3180 [mdt]
[11121.494617] [<0>] mdt_reint_rec+0x117/0x270 [mdt]
[11121.495442] [<0>] mdt_reint_internal+0x4bf/0x7d0 [mdt]
[11121.496318] [<0>] mdt_intent_open+0x137/0x420 [mdt]
[11121.497164] [<0>] mdt_intent_opc+0x12c/0xbf0 [mdt]
[11121.497980] [<0>] mdt_intent_policy+0x207/0x3a0 [mdt]
[11121.499224] [<0>] ldlm_lock_enqueue+0x47b/0xb20 [ptlrpc]
[11121.500246] [<0>] ldlm_handle_enqueue0+0x634/0x1520 [ptlrpc]
[11121.501234] [<0>] tgt_enqueue+0xa4/0x210 [ptlrpc]
[11121.502079] [<0>] tgt_request_handle+0xcbf/0x1920 [ptlrpc]
[11121.503032] [<0>] handle_recovery_req+0x140/0x270 [ptlrpc]
[11121.504009] [<0>] replay_request_or_update.isra.31+0x2fa/0xa20 [ptlrpc]
[11121.505135] [<0>] target_recovery_thread+0x736/0x1320 [ptlrpc]
[11121.506146] [<0>] kthread+0x116/0x130
[11121.506798] [<0>] ret_from_fork+0x35/0x40

A comment of Dongyang Li in ~~LU-14692~~ say that this could be linked to https://review.whamcloud.com/c/fs/lustre-release/+/38424/ ("~~LU-11912~~ ofd: reduce LUSTRE_DATA_SEQ_MAX_WIDTH").

Whamcloud Community JIRA

[LU-16692] replay-single: test_70c osp_fid_diff()) ASSERTION( fid_seq(fid1) == fid_seq(fid2) )