[LU-10843] snapshot_create fails after MGS restart

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.10.3
• Fix Version/s: Lustre 2.12.0, Lustre 2.10.5
• Type: Bug
• Priority: Major
• Reporter: Nathaniel Clark
• Assignee: Nathaniel Clark
• Resolution: Fixed
• Severity: 3

Description

If the MGS is unmounted and remount (or failed over) lctl snapshot_create fails with the following error:

# lctl snapshot_create -F scratch -n test
Fail to freeze barrier for scratch: No such device
Can't create the snapshot test

snapshot_destroy still works without issue.

WORKAROUND: unmount/mount the MDT and snapshot_create will commence working again.

NOTE: mount -o remount is insufficient

Comments

[Nathaniel Clark]

After remount of MGS FSNAME-barrier is missing from mgs->mgs_fs_db_list. It is only added during mgs_handler.c::mgs_target_reg()

[Gerrit Updater]

Nathaniel Clark ([e-mail]) uploaded a new patch: https://review.whamcloud.com/31779
Subject: LU-10843 mgs: allow snapshot after MGS remount
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: d4487c2e5cdc8ac4e0f16b29e9bd05ee3bd60edb

[Nathaniel Clark]

In debugging #31779 I used wireshare to figure out when the MGS actually started accepting barrier commands and to that end, I've added barrier parsing to wireshark:

https://code.wireshark.org/review/#/c/26692/

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/31779/
Subject: LU-10843 mgs: allow snapshot after MGS remount
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 2d63d5bc74a5b2477d795bf9e9b0eba94f7fc19f

[Peter Jones]

Landed for 2.12

[Gerrit Updater]

Minh Diep ([e-mail]) uploaded a new patch: https://review.whamcloud.com/32519
Subject: LU-10843 mgs: allow snapshot after MGS remount
Project: fs/lustre-release
Branch: b2_10
Current Patch Set: 1
Commit: 02c6dc424eea35cb79840f2dd1144543a502ccd9

[Gerrit Updater]

John L. Hammond ([e-mail]) merged in patch https://review.whamcloud.com/32519/
Subject: LU-10843 mgs: allow snapshot after MGS remount
Project: fs/lustre-release
Branch: b2_10
Current Patch Set:
Commit: f01f2174bcc27c06c8cd5e64cb6e63452fe6c07b