From: Chris Murphy <lists@colorremedies.com>
To: Gregory Petit <gregory@amphorawinery.eu>,
Btrfs BTRFS <linux-btrfs@vger.kernel.org>
Subject: Re: RAID10: uncorrectable errors
Date: Thu, 12 Jan 2017 23:46:44 -0700 [thread overview]
Message-ID: <CAJCQCtTw-tUA5B0GD9i4WOncZ5DO=Nkf+71oZX23Mpw9zN4hdA@mail.gmail.com> (raw)
In-Reply-To: <707f7f482dd48c08544574b71ab77299@amphorawinery.eu>
On Thu, Jan 12, 2017, 2:55 AM Gregory Petit <gregory@amphorawinery.eu> wrote:
> Here are the details:
>
> scrub started at Wed Jan 11 18:00:01 2017 and finished after 00:19:23
> total bytes scrubbed: 1.14TiB with 4 errors
> error details: csum=4
> corrected errors: 0, uncorrectable errors: 4, unverified errors: 0
>
> From dmesg:
> Wed Jan 11 18:10:35 2017] BTRFS error (device sda): bdev /dev/sda errs:
> wr 0, rd 0, flush 0, corrupt 1, gen 0
> [Wed Jan 11 18:10:35 2017] BTRFS error (device sda): unable to fixup
> (regular) error at logical 631657844736 on dev /dev/sda
> [Wed Jan 11 18:10:51 2017] BTRFS error (device sda): bdev /dev/sdb errs:
> wr 0, rd 0, flush 0, corrupt 1, gen 0
> [Wed Jan 11 18:10:51 2017] BTRFS error (device sda): unable to fixup
> (regular) error at logical 632954847232 on dev /dev/sdb
> [Wed Jan 11 18:18:57 2017] BTRFS error (device sda): bdev /dev/sdc errs:
> wr 0, rd 0, flush 0, corrupt 1, gen 0
> [Wed Jan 11 18:18:57 2017] BTRFS error (device sda): unable to fixup
> (regular) error at logical 632954847232 on dev /dev/sdc
> [Wed Jan 11 18:19:19 2017] BTRFS error (device sda): bdev /dev/sde errs:
> wr 0, rd 0, flush 0, corrupt 1, gen 0
> [Wed Jan 11 18:19:19 2017] BTRFS error (device sda): unable to fixup
> (regular) error at logical 631657844736 on dev /dev/sde
Look at the logical addresses. Two pair, four total, have errors.
Looks like both copies of two blocks of information are corrupt, and
that's why fix up doesn't happen. I'm gonna guess this is metadata.
But between 'btrfs inspect-internal logical-resolve' or 'dump-tree'
with those two block numbers, you should be able to figure out what's
affected. Pretty strange for both copies to get munged though, but I'm
suspicious of hardware - in particular controller or cable or even
RAM, since it affects at least two drives. The chances this is two
drives corrupting the same logical block of data is almost zero.
>
> root@proxmox:~# btrfs check --repair /dev/sda
FWIW btrfs check finds all member devices for you regardless of which
device you point it to, and checks the whole file system. It's not
necessary to run it on each device.
'btrfs check --mode=lowmem' might find the problem but I don't think
it can fix anything still.
Chris Murphy
prev parent reply other threads:[~2017-01-13 6:46 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-12 9:45 RAID10: uncorrectable errors Gregory Petit
2017-01-13 6:46 ` Chris Murphy [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAJCQCtTw-tUA5B0GD9i4WOncZ5DO=Nkf+71oZX23Mpw9zN4hdA@mail.gmail.com' \
--to=lists@colorremedies.com \
--cc=gregory@amphorawinery.eu \
--cc=linux-btrfs@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).