Announcing: ZFS-fuse 0.7.0
#69 — zfs clear not working
| State | Unconfirmed |
|---|---|
| Version: | 0.7.0 |
| Area | Functionality |
| Issue type | Bug |
| Severity | Medium |
| Submitted by | Jan Ploski |
| Submitted on | Jul 04, 2010 |
| Responsible | Seth Heeren |
| Target release: |
—
|
Last modified on
Sep 20, 2010
by
Jan Ploski
I'm using snapshot sehe-288ab55443945461f8f8fe02221b37aafa9557cf. I had some files that have been reported as corrupted (and could not be read). I removed these files and wanted to get rid of the error messages in the pool. But "zfs clear" doesn't do anything:
remotejava:~# zpool clear green
remotejava:~# zpool status green -v
pool: green
state: ONLINE
status: One or more devices has experienced an error resulting in data
corruption. Applications may be affected.
action: Restore the file in question if possible. Otherwise restore the
entire pool from backup.
see: http://www.sun.com/msg/ZFS-8000-8A
scrub: none requested
config:
NAME STATE READ WRITE CKSUM
green ONLINE 0 0 0
disk/by-id/dm-name-green ONLINE 0 0 0
errors: Permanent errors have been detected in the following files:
<metadata>:<0x0>
<0x21f>:<0x4e>
<0x21f>:<0x53>
<0x21f>:<0x5a>
<0x21f>:<0x62>
<0x21f>:<0x68>
green/backup/vista:<0x4e>
green/backup/vista:<0x53>
green/backup/vista:<0x67>
green/backup/vista:<0x68>
remotejava:~# zpool clear green
remotejava:~# zpool status green -v
pool: green
state: ONLINE
status: One or more devices has experienced an error resulting in data
corruption. Applications may be affected.
action: Restore the file in question if possible. Otherwise restore the
entire pool from backup.
see: http://www.sun.com/msg/ZFS-8000-8A
scrub: none requested
config:
NAME STATE READ WRITE CKSUM
green ONLINE 0 0 0
disk/by-id/dm-name-green ONLINE 0 0 0
errors: Permanent errors have been detected in the following files:
<metadata>:<0x0>
<0x21f>:<0x4e>
<0x21f>:<0x53>
<0x21f>:<0x5a>
<0x21f>:<0x62>
<0x21f>:<0x68>
green/backup/vista:<0x4e>
green/backup/vista:<0x53>
green/backup/vista:<0x67>
green/backup/vista:<0x68>
Added by
(anonymous)
on
Jul 04, 2010 06:20 PM
erm... doesn't look good
it is my understanding that you need to remove the failed files; the status message will then transform into a different error report that (I think?) explicitely says that you can clear the message using zpool clear.
I'm missing this bit in the text right now.
I'm a bit worried that it lists meta-data problems because, as I understand it, these cannot be recovered from (although you might have luck destroying the related filesystems if at all possible?)
On error recovery, I'm in no way an expert, and I usually refer to upstream (zfs-discuss, e.g.)
it is my understanding that you need to remove the failed files; the status message will then transform into a different error report that (I think?) explicitely says that you can clear the message using zpool clear.
I'm missing this bit in the text right now.
I'm a bit worried that it lists meta-data problems because, as I understand it, these cannot be recovered from (although you might have luck destroying the related filesystems if at all possible?)
On error recovery, I'm in no way an expert, and I usually refer to upstream (zfs-discuss, e.g.)
Added by
Jan Ploski
on
Jul 04, 2010 07:05 PM
The last four entries (and perhaps the first few as well?) used to contain file names. After I deleted the files, the file names disappeared from the error messages and were replaced by hex codes. I'm not sure, but I believe that after destroying the dataset, the errors concerning 'green/backup/vista' disappeared altogether, but now after recreating it and copying back files I again have a few newer ones.
It's a weak clue, but all the errors I encountered in my pool so far are all triggered by rsyncing (system) files from Vista, several other datasets that are used to rsync from Linux boxes are all fine (so far).
It's a weak clue, but all the errors I encountered in my pool so far are all triggered by rsyncing (system) files from Vista, several other datasets that are used to rsync from Linux boxes are all fine (so far).
Added by
Seth Heeren
on
Sep 19, 2010 06:23 PM
Responsible manager:
(UNASSIGNED) → sgheeren
> It's a weak clue, but all the errors I encountered in my pool so far are all triggered by rsyncing (system) files from Vista, several other datasets that are used to rsync from Linux boxes are all fine (so far).
Does rsync do anything with acls? In that case the patch in #71 might alleviate things. It'll be in 0.7.0 and is available in testing
Is this item ok for closing, or is this still causing problems for you?
Does rsync do anything with acls? In that case the patch in #71 might alleviate things. It'll be in 0.7.0 and is available in testing
Is this item ok for closing, or is this still causing problems for you?
Added by
Jan Ploski
on
Sep 20, 2010 08:18 AM
Well, it's hard to tell whether this issue is resolved. I still have a bunch of nonsensical "errors" in my pool from this old episode, which zpool clear cannot remove. The hex codes are now gone and instead some paths (referring to directories) appear. I can access the contents of those directories normally. I didn't have time nor motivation to copy away all the files, recreate the pool from scratch, and then restore the files just to get rid of these spurious errors. This doesn't seem like a satisfactory solution. On the other hand, no new errors of this sort have appeared since my original report, so I'd say that the "patient's condition is stable".
