Do not set MD_BROKEN on failfast io failure #240
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
![kpd-daemon[bot]](https://avatars.githubusercontent.com/u/162645851?s=80&v=4){kind=small}
This commit ensures that an MD_FAILFAST IO failure does not put
the array into a broken state.
When failfast is enabled on rdev in RAID1 or RAID10,
the array may be flagged MD_BROKEN in the following cases.
- If MD_FAILFAST IOs to multiple rdevs fail simultaneously
- If an MD_FAILFAST metadata write to the 'last' rdev fails
The MD_FAILFAST bio error handler always calls md_error on IO failure,
under the assumption that raid{1,10}_error will neither fail
the last rdev nor break the array.
After commit 9631abd ("md: Set MD_BROKEN for RAID1 and RAID10"),
calling md_error on the 'last' rdev in RAID1/10 always sets
the MD_BROKEN flag on the array.
As a result, when failfast IO fails on the last rdev, the array
immediately becomes failed.
Normally, MD_FAILFAST IOs are not issued to the 'last' rdev, so this is
an edge case; however, it can occur if rdevs in a non-degraded
array share the same path and that path is lost, or if a metadata
write is triggered in a degraded array and fails due to failfast.
When a failfast metadata write fails, it is retried through the
following sequence. If a metadata write without failfast fails,
the array will be marked with MD_BROKEN.
1. MD_SB_NEED_REWRITE is set in sb_flags by super_written.
2. md_super_wait, called by the function executing md_super_write,
returns -EAGAIN due to MD_SB_NEED_REWRITE.
3. The caller of md_super_wait (e.g., md_update_sb) receives the
negative return value and retries md_super_write.
4. md_super_write issues the metadata write again,
this time without MD_FAILFAST.
Fixes: 9631abd ("md: Set MD_BROKEN for RAID1 and RAID10")
Signed-off-by: Kenta Akagi <[email protected]>
Once MD_BROKEN is set on an array, no further writes can be performed to it. The user must be informed that the array cannot continue operation. Signed-off-by: Kenta Akagi <[email protected]>
Since commit 9a56784 ("md: allow last device to be forcibly removed from RAID1/RAID10."), RAID1/10 arrays can now lose all rdevs. Before that commit, losing the array last rdev or reaching the end of the function without early return in raid{1,10}_error never occurred. However, both situations can occur in the current implementation. As a result, when mddev->fail_last_dev is set, a spurious pr_crit message can be printed. This patch prevents "Operation continuing" printed if the array is not operational. root@fedora:~# mdadm --create --verbose /dev/md0 --level=1 \ --raid-devices=2 /dev/loop0 /dev/loop1 mdadm: Note: this array has metadata at the start and may not be suitable as a boot device. If you plan to store '/boot' on this device please ensure that your boot-loader understands md/v1.x metadata, or use --metadata=0.90 mdadm: size set to 1046528K Continue creating array? y mdadm: Defaulting to version 1.2 metadata mdadm: array /dev/md0 started. root@fedora:~# echo 1 > /sys/block/md0/md/fail_last_dev root@fedora:~# mdadm --fail /dev/md0 loop0 mdadm: set loop0 faulty in /dev/md0 root@fedora:~# mdadm --fail /dev/md0 loop1 mdadm: set device faulty failed for loop1: Device or resource busy root@fedora:~# dmesg | tail -n 4 [ 1314.359674] md/raid1:md0: Disk failure on loop0, disabling device. md/raid1:md0: Operation continuing on 1 devices. [ 1315.506633] md/raid1:md0: Disk failure on loop1, disabling device. md/raid1:md0: Operation continuing on 0 devices. root@fedora:~# Fixes: 9a56784 ("md: allow last device to be forcibly removed from RAID1/RAID10.") Signed-off-by: Kenta Akagi <[email protected]>
|
Upstream branch: c17fb54 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Pull request for series with
subject: Do not set MD_BROKEN on failfast io failure
version: 3
url: https://patchwork.kernel.org/project/linux-raid/list/?series=996574