[LU-6276] LustreError: error destroying object

Duplicate — Wed, 25 Feb 2015 18:18:43 +0000

I see the following LustreErrors showing up in irregular intervals:

Feb 24 15:03:47 oss-10 kernel: LustreError: 2417:0:(ofd_obd.c:1096:ofd_destroy()) vegas-OST0052: error destroying object [0x100000000:0x1d0e5:0x0]: 0
Feb 24 15:03:47 oss-10 kernel: LustreError: 2417:0:(ofd_obd.c:1096:ofd_destroy()) Skipped 300 previous similar messages
Feb 24 15:04:53 oss-10 kernel: Lustre: vegas-OST0052-os: trigger OI scrub by RPC for [0x100000000:0xc4a6:0x0], rc = 0 [1]
Feb 24 15:04:53 oss-10 kernel: Lustre: Skipped 219 previous similar messages
Feb 24 15:13:51 oss-10 kernel: LustreError: 2417:0:(ofd_obd.c:1096:ofd_destroy()) vegas-OST0052: error destroying object [0x100000000:0x20c7:0x0]: 0
Feb 24 15:13:51 oss-10 kernel: LustreError: 2417:0:(ofd_obd.c:1096:ofd_destroy()) Skipped 300 previous similar messages
Feb 24 15:14:57 oss-10 kernel: Lustre: vegas-OST0052-os: trigger OI scrub by RPC for [0x100000000:0x20c7:0x0], rc = 0 [1]
Feb 24 15:14:57 oss-10 kernel: Lustre: Skipped 215 previous similar messages

The object ID is not changing. It is always one of those two mentioned. This is only happening on one OST (out of 90). Filesystem seems to be fine otherwise.

output of lctl get_param -n osd-ldiskfs.vegas-OST0052.oi_scrub
name: OI_scrub
magic: 0x4c5fd252
oi_files: 64
status: completed
flags:
param:
time_since_last_completed: 4 seconds
time_since_latest_start: 4 seconds
time_since_last_checkpoint: 4 seconds
latest_start_position: 12
last_checkpoint_position: 40054145
first_failure_position: N/A
checked: 43262
updated: 0
failed: 0
prior_updated: 0
noscrub: 0
igif: 2
success_count: 10871
run_time: 0 seconds
average_speed: 43262 objects/sec
real-time_speed: N/A
current_position: N/A
lf_scanned: 0
lf_reparied: 0
lf_failed: 0

The OI_scrub seems to restart every 10 seconds or so?

What is the problem and is there a cure?

Whamcloud Community JIRA

[LU-6276] LustreError: error destroying object