[LU-16873] LU-16655 fix doesn't update OI_Scrub with new magic Created: 05/Jun/23  Updated: 02/Feb/24  Resolved: 14/Jun/23

Status: Resolved
Project: Lustre
Component/s: None
Affects Version/s: Lustre 2.15.3
Fix Version/s: Lustre 2.16.0, Lustre 2.15.4

Type: Bug Priority: Major
Reporter: Alexander Zarochentsev Assignee: Alexander Zarochentsev
Resolution: Fixed Votes: 0
Labels: None

Issue Links:
Related
is related to LU-16655 Files not accessible after 2.12 -> 2.... Resolved
Severity: 3
Rank (Obsolete): 9223372036854775807

 Description   

The fix for LU-16655 detects the format change correctly but does not write new oi scrub file magic , so new mount triggers the "oi files counter reset" again and again:

[root@devvm4 tests]# LCTL=/home/zam/work/git/lustre-wc-rel/lustre/utils/lctl
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t ldiskfs /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# dd if=$HOME/OI_scrub.v1 of=/mnt/lustre-mds1/OI_scrub seek=16 skip=16 bs=1 conv=notrunc
384+0 records in
384+0 records out
384 bytes copied, 0.00154305 s, 249 kB/s
[root@devvm4 tests]# echo -n -e '\x01\x00' | dd of=/mnt/lustre-mds1/OI_scrub bs=1 count=2 status=none seek=136 conv=notrunc 
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t lustre /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# $LCTL get_param *.*.oi_scrub | grep oi_files
oi_files: 64
oi_files: 64
oi_files: 64
[root@devvm4 tests]# 
[root@devvm4 tests]# dmesg  | grep LU-16655
[ 3325.250365] Lustre: 62237:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t lustre /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# dmesg  | grep LU-16655
[ 3325.250365] Lustre: 62237:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[ 3344.469841] Lustre: 62344:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t lustre /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# dmesg  | grep LU-16655
[ 3325.250365] Lustre: 62237:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[ 3344.469841] Lustre: 62344:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t lustre /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# dmesg  | grep LU-16655
[ 3325.250365] Lustre: 62237:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[ 3344.469841] Lustre: 62344:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[ 3379.617818] Lustre: 62529:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t lustre /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# dmesg  | grep LU-16655
[ 3325.250365] Lustre: 62237:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[ 3344.469841] Lustre: 62344:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[ 3379.617818] Lustre: 62529:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[ 3432.251449] Lustre: 62622:0:(scrub.c:189:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[root@devvm4 tests]#


 Comments   
Comment by Gerrit Updater [ 05/Jun/23 ]

"Alexander Zarochentsev <alexander.zarochentsev@hpe.com>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/51226
Subject: LU-16873 osd: update OI_Scrub file with new magic
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 255d80005018ab51dd41c6d8be60745c68194450

Comment by Alexander Zarochentsev [ 05/Jun/23 ]

the test from the description after applying the fix:

...
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t lustre /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# dmesg | grep LU-16655
[ 3941.062617] Lustre: 68735:0:(scrub.c:191:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[root@devvm4 tests]# umount /mnt/lustre-mds1/
[root@devvm4 tests]# mount -t lustre /dev/mapper/mds1_flakey /mnt/lustre-mds1/
[root@devvm4 tests]# dmesg | grep LU-16655
[ 3941.062617] Lustre: 68735:0:(scrub.c:191:scrub_file_load()) lustre-MDT0000: reset scrub OI count for format change (LU-16655)
[root@devvm4 tests]#
Comment by Andreas Dilger [ 05/Jun/23 ]

Zam, just to confirm, the LU-16655 fix still works correctly, and the console message is confusing/annoying, but not a sign of problems? Just trying to determine if this is a must-have patch for the 2.15.3 release, or if it is not critical (and more important to get the actual LU-16655 fix released).

Comment by Alexander Zarochentsev [ 06/Jun/23 ]

adilger ,
>Zam, just to confirm, the LU-16655 fix still works correctly, and the console message is confusing/annoying, but not a sign of problems?

it doesn't break anything. it only recalculates the number of oi files each mount and prints the error message.

Comment by Gerrit Updater [ 14/Jun/23 ]

"Oleg Drokin <green@whamcloud.com>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/51226/
Subject: LU-16873 osd: update OI_Scrub file with new magic
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 38b7c408212f60d684c9b114d90b4514e0044ffe

Comment by Peter Jones [ 14/Jun/23 ]

Landed for 2.16

Comment by Gerrit Updater [ 30/Jun/23 ]

"Etienne AUJAMES <eaujames@ddn.com>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/51525
Subject: LU-16873 osd: update OI_Scrub file with new magic
Project: fs/lustre-release
Branch: b2_15
Current Patch Set: 1
Commit: b31c29a3611c02d4a8e9f9c5023f4a7497cff7dc

Comment by Gerrit Updater [ 19/Oct/23 ]

"Oleg Drokin <green@whamcloud.com>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/51525/
Subject: LU-16873 osd: update OI_Scrub file with new magic
Project: fs/lustre-release
Branch: b2_15
Current Patch Set:
Commit: 703d8a994bf33ac07f8bd4c956880db3d9abb016

Generated at Sat Feb 10 03:30:43 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.