backref mismatch / backpointer mismatch

backref mismatch / backpointer mismatch

[Johannes Rohr]

Dear all,

Last night I did a readonly offline check of our server's file system
and I got a bunch of backref mismatch / backpointer mismatch errors.

Since everybody is warning NOT to use btrfsck --repair, I wonder what
options I have of fixing this issue, short of reformatting the whole disk

All the info (uname -a, btrfs fi usage, btrfs check output) is here:

https://gist.github.com/vasyugan/45471206df5d9981f51b9bbf9362508f

Thanks so much in advance for your advice,

Johannes

Re: backref mismatch / backpointer mismatch

[Qu Wenruo]

[-- Attachment #1.1: Type: text/plain, Size: 726 bytes --]



On 2020/9/9 下午10:51, Johannes Rohr wrote:
> Dear all,
> 
> Last night I did a readonly offline check of our server's file system
> and I got a bunch of backref mismatch / backpointer mismatch errors.
> 
> Since everybody is warning NOT to use btrfsck --repair, I wonder what
> options I have of fixing this issue, short of reformatting the whole disk
> 
> All the info (uname -a, btrfs fi usage, btrfs check output) is here:
> 
> https://gist.github.com/vasyugan/45471206df5d9981f51b9bbf9362508f

What about `btrfs check --mode=lowmem` result?

If your btrfs-progs is not uptodate, it may be a false alert.

Thanks,
Qu
> 
> Thanks so much in advance for your advice,
> 
> Johannes
> 
> 
> 


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: backref mismatch / backpointer mismatch

[Johannes Rohr]

[-- Attachment #1.1: Type: text/plain, Size: 1087 bytes --]

Am 10.09.20 um 04:16 schrieb Qu Wenruo:
>
> On 2020/9/9 下午10:51, Johannes Rohr wrote:

> What about `btrfs check --mode=lowmem` result?

Too late already. After getting the green light in #btrfs, I ran the
--repair and it fixed the issue.

>
> If your btrfs-progs is not uptodate, it may be a false alert.

The reason why I did the offline check in the first place was that there
were issues, namely, that btrfs balance start -musage=70 / resulted in a
segfault and also my attempts to remove a faulty sdd from the RAID
reproducibly led to a kernel oops and a hung filesystem. I had asked
about this two times on the list [1] also reported the issue to the
kernel bugzilla [2]. Also,  upgraded btrfs-progs to 5.7 before running
the test.

Fortunately --repair has fixed these issues.

Cheers,

Johannes


[1] https://www.mail-archive.com/linux-btrfs@vger.kernel.org/msg88916.html

[2] , https://bugzilla.kernel.org/show_bug.cgi?id=209143




>
> Thanks,
> Qu



>> Thanks so much in advance for your advice,
>>
>> Johannes
>>
>>
>>



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]