openzfs/zfs

Resilver restarts repeatedly when checkpoint is present

Open

#17,109 opened on 2025年3月3日

GitHub で見る
 (1 comment) (1 reaction) (0 assignees)C (9,908 stars) (1,703 forks)batch import
Type: Defectgood first issue

説明

System information

Type Version/Name
Distribution Name NixOS
Distribution Version unstable
Kernel Version 6.6
Architecture x86_64
OpenZFS Version 2.2.7

Describe the problem you're observing

Resilver restarts repeatedly when checkpoint is present. Discarding the checkpoint resolves.

Describe how to reproduce the problem

I do not have specific, repeatably reproduction steps at this time. Vaguely, though:

  1. zpool checkpoint tank
  2. Cause a disk to need a resilver
  3. zpool clear the disk
  4. …clear the checkpoint to fix resilver constantly restarting

Include any warning/errors/backtraces from the system logs

This is truncated.

Mar  3 2025 00:08:23.950317618 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "starting deferred resilver"
        history_txg = 0x9620f1
        history_time = 0x67c4f2f7
        time = 0x67c4f2f7 0x38a4b232 
        eid = 0x455

Mar  3 2025 00:08:29.172279886 sysevent.fs.zfs.resilver_start
        version = 0x0
        class = "sysevent.fs.zfs.resilver_start"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        resilver_type = "healing"
        time = 0x67c4f2fd 0xa44c84e 
        eid = 0x456

Mar  3 2025 00:08:29.172279886 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "func=2 mintxg=9782617 maxtxg=9838805"
        history_internal_name = "scan setup"
        history_txg = 0x9620f2
        history_time = 0x67c4f2fd
        time = 0x67c4f2fd 0xa44c84e 
        eid = 0x457

Mar  3 2025 00:09:02.500040402 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "scan done"
        history_txg = 0x9620f9
        history_time = 0x67c4f31e
        time = 0x67c4f31e 0x1dce02d2 
        eid = 0x458

Mar  3 2025 00:09:02.501040395 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "starting deferred resilver"
        history_txg = 0x9620f9
        history_time = 0x67c4f31e
        time = 0x67c4f31e 0x1ddd450b 
        eid = 0x459

Mar  3 2025 00:09:07.572004153 sysevent.fs.zfs.resilver_start
        version = 0x0
        class = "sysevent.fs.zfs.resilver_start"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        resilver_type = "healing"
        time = 0x67c4f323 0x22181739 
        eid = 0x45a

Mar  3 2025 00:09:07.572004153 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "func=2 mintxg=9782617 maxtxg=9838805"
        history_internal_name = "scan setup"
        history_txg = 0x9620fa
        history_time = 0x67c4f323
        time = 0x67c4f323 0x22181739 
        eid = 0x45b

Mar  3 2025 00:09:39.641776094 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "scan done"
        history_txg = 0x962101
        history_time = 0x67c4f343
        time = 0x67c4f343 0x2640b9de 
        eid = 0x45c

Mar  3 2025 00:09:39.642776087 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "starting deferred resilver"
        history_txg = 0x962101
        history_time = 0x67c4f343
        time = 0x67c4f343 0x264ffc17 
        eid = 0x45d

Mar  3 2025 00:09:44.947738547 sysevent.fs.zfs.resilver_start
        version = 0x0
        class = "sysevent.fs.zfs.resilver_start"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        resilver_type = "healing"
        time = 0x67c4f348 0x387d57b3 
        eid = 0x45e

Mar  3 2025 00:09:44.947738547 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "func=2 mintxg=9782617 maxtxg=9838805"
        history_internal_name = "scan setup"
        history_txg = 0x962102
        history_time = 0x67c4f348
        time = 0x67c4f348 0x387d57b3 
        eid = 0x45f

Mar  3 2025 00:10:16.871513707 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "scan done"
        history_txg = 0x962109
        history_time = 0x67c4f368
        time = 0x67c4f368 0x33f23e6b 
        eid = 0x460

Mar  3 2025 00:10:16.872513700 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "starting deferred resilver"
        history_txg = 0x962109
        history_time = 0x67c4f368
        time = 0x67c4f368 0x340180a4 
        eid = 0x461

Mar  3 2025 00:10:22.323475486 sysevent.fs.zfs.resilver_start
        version = 0x0
        class = "sysevent.fs.zfs.resilver_start"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        resilver_type = "healing"
        time = 0x67c4f36e 0x1347d81e 
        eid = 0x462

Mar  3 2025 00:10:22.323475486 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "func=2 mintxg=9782617 maxtxg=9838805"
        history_internal_name = "scan setup"
        history_txg = 0x96210a
        history_time = 0x67c4f36e
        time = 0x67c4f36e 0x1347d81e 
        eid = 0x463

Mar  3 2025 00:10:40.365349358 resource.fs.zfs.statechange
        version = 0x0
        class = "resource.fs.zfs.statechange"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        vdev_guid = 0xb4ec113378d257d1
        vdev_state = "FAULTED" (0x5)
        vdev_path = "/dev/mydisk"
        vdev_laststate = "ONLINE" (0x7)
        time = 0x67c4f380 0x15c6c9ee 
        eid = 0x464

Mar  3 2025 00:10:50.011282144 sysevent.fs.zfs.config_sync
        version = 0x0
        class = "sysevent.fs.zfs.config_sync"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        time = 0x67c4f38a 0xac26e0 
        eid = 0x465

Mar  3 2025 00:10:52.673263622 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "scan done"
        history_txg = 0x962112
        history_time = 0x67c4f38c
        time = 0x67c4f38c 0x28213006 
        eid = 0x466

Mar  3 2025 00:11:25.640035151 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "started discarding checkpointed state from the pool"
        history_internal_name = "spa discard checkpoint"
        history_txg = 0x962119
        history_time = 0x67c4f3ad
        time = 0x67c4f3ad 0x2626294f 
        eid = 0x467

Mar  3 2025 00:11:26.058032265 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "finished discarding checkpointed state from the pool"
        history_internal_name = "spa discard checkpoint"
        history_txg = 0x96211b
        history_time = 0x67c4f3ae
        time = 0x67c4f3ae 0x3758089 
        eid = 0x468

Mar  3 2025 00:11:32.473987996 resource.fs.zfs.statechange
        version = 0x0
        class = "resource.fs.zfs.statechange"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        vdev_guid = 0xb4ec113378d257d1
        vdev_state = "ONLINE" (0x7)
        vdev_path = "/dev/mydisk"
        vdev_laststate = "FAULTED" (0x5)
        time = 0x67c4f3b4 0x1c407b9c 
        eid = 0x469

Mar  3 2025 00:11:32.580987259 sysevent.fs.zfs.vdev_clear
        version = 0x0
        class = "sysevent.fs.zfs.vdev_clear"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        vdev_guid = 0xb4ec113378d257d1
        vdev_state = "ONLINE" (0x7)
        vdev_path = "/dev/mydisk"
        time = 0x67c4f3b4 0x22a1297b 
        eid = 0x46a

Mar  3 2025 00:11:32.864985301 sysevent.fs.zfs.resilver_start
        version = 0x0
        class = "sysevent.fs.zfs.resilver_start"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        resilver_type = "healing"
        time = 0x67c4f3b4 0x338ea0d5 
        eid = 0x46b

Mar  3 2025 00:11:32.864985301 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "func=2 mintxg=9782617 maxtxg=9838877"
        history_internal_name = "scan setup"
        history_txg = 0x96211e
        history_time = 0x67c4f3b4
        time = 0x67c4f3b4 0x338ea0d5 
        eid = 0x46c

Mar  3 2025 00:11:39.313940869 sysevent.fs.zfs.config_sync
        version = 0x0
        class = "sysevent.fs.zfs.config_sync"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        time = 0x67c4f3bb 0x12b65b85 
        eid = 0x46d

Mar  3 2025 00:12:05.063764056 sysevent.fs.zfs.history_event
        version = 0x0
        class = "sysevent.fs.zfs.history_event"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        history_hostname = "myhostname"
        history_internal_str = "errors=0"
        history_internal_name = "scan done"
        history_txg = 0x962126
        history_time = 0x67c4f3d5
        time = 0x67c4f3d5 0x3ccf658 
        eid = 0x46e

Mar  3 2025 00:12:05.063764056 sysevent.fs.zfs.resilver_finish
        version = 0x0
        class = "sysevent.fs.zfs.resilver_finish"
        pool = "mytank"
        pool_guid = 0xec15e7c2130c98d4
        pool_state = 0x0
        pool_context = 0x0
        resilver_type = "healing"
        time = 0x67c4f3d5 0x3ccf658 
        eid = 0x46f

コントリビューターガイド

技術スタック
c
領域
backend
Issue 種別
bug
難度新しい貢献者にとっての実装難度です。1 は非常に小さな変更、5 は専門的な作業です。
5
推定時間経験ある貢献者が調査、実装、テスト、pull request 準備にかけるおおよその時間範囲です。
over 1 week
活動状況Issue が今どれだけ取り組みやすいかを示します。新しい、活発、古い、ブロック中、またはメンテナー入力待ちなどです。
fresh
明確さ期待される変更、受け入れ条件、次の手順がどれだけ明確かを示します。
needs investigation
前提条件
C programmingZFS internalsCheckpoint mechanism
初心者向け度初回貢献者にどれだけ取り組みやすいかを 1-100 で推定したスコアです。
15
調査方針
Investigate the interaction between checkpoint and resilver in the ZFS codebase. Focus on the scan state machine in module/zfs/dsl scan.c and the checkpoint routines in module/zfs/spa checkpoint.c. Reproduce the issue by setting up a pool with a checkpoint and triggering a resilver. Look for conditions where checkpoint presence causes scan restart logic to loop. See also existing tests for checkpoint and resilver in tests/zfs tests/tests/.