linux-bcache.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Constantin Runge <c.runge@cssbook.de>
To: linux-bcache@vger.kernel.org
Subject: Obsoleted large bucket layout message once
Date: Tue, 16 Feb 2021 16:06:41 +0100	[thread overview]
Message-ID: <de528a90-6d3d-d324-ca22-73b635d5153f@cssbook.de> (raw)

Hi everybody,

when registering and mounting my bcache devices (unortunately my root 
filesystem), the kernel informed me about `Detected obsoleted large 
bucket layout, all attached bcache device will be read-only'.

Subsequently, my root fs was, in fact, read-only. Thus, I have no logs 
for you unfortunately. As far as I remember, bcache-super-show showed 
the same bucket and block sizes as always (except that the backing 
device showed a clean cache state):

 > bcache-super-show /dev/sda6
sb.magic		ok
sb.first_sector		8 [match]
sb.csum			C5EB3BA8CA523D61 [match]
sb.version		4 [backing device]

dev.label		(empty)
dev.uuid		264d1acc-8461-4ddc-a825-f0acdeea3ede
dev.sectors_per_block	1
dev.sectors_per_bucket	1024
dev.data.first_sector	2048
dev.data.cache_mode	1 [writeback]
dev.data.cache_state	2 [dirty]

cset.uuid		755901be-a101-4581-b1d0-4d21d1619f27
 > bcache-super-show /dev/sdb3
sb.magic		ok
sb.first_sector		8 [match]
sb.csum			BDA92F1E72DBA095 [match]
sb.version		3 [cache device]

dev.label		(empty)
dev.uuid		25151c1a-904e-492e-b99b-6d07e13b047a
dev.sectors_per_block	1
dev.sectors_per_bucket	1024
dev.cache.first_sector	1024
dev.cache.cache_sectors	907564032
dev.cache.total_sectors	907565056
dev.cache.ordered	yes
dev.cache.discard	no
dev.cache.pos		0
dev.cache.replacement	0 [lru]

cset.uuid		755901be-a101-4581-b1d0-4d21d1619f27


For debugging purposes I tried to mount and chroot into my backing 
device directly, using a loopback interface. I was not able to find an 
offset for `losetup -f -o <offset> /dev/sda6', so that `mount -o loop 
/dev/loop0 /mnt/mnt0' would not tell me that there was no filesystem.
Also, `for i in {0.2000}; do dd if=/dev/sda6 skip=$i | file -; done' 
from [1] showed `data' for every offset.

After getting a little desperate, I booted into an older kernel 
(although I didn't update my kernel or my bcache-tools since the last 
boot) and everything worked. I then booted back into my current system 
and everything worked again.

Is this a bug with bcache or do I need to change my setup to prevent 
this from happening again? My bcache devices were created sometime 
around 2015 or 2016.

System info:
uname -a: `Linux Horus 5.10.10-gentoo #1 SMP PREEMPT Mon Feb 15 15:00:49 
CET 2021 x86_64 Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz GenuineIntel 
GNU/Linux'
bcache-tools version: 1.1
The older kernel is a 5.8.11-gentoo with voluntary preempt.

Best regards and thanks,
Constantin


   [1] 
https://stackoverflow.com/questions/22820492/how-to-revert-bcache-device-to-regular-device

             reply	other threads:[~2021-02-16 15:14 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-16 15:06 Constantin Runge [this message]
2021-02-18  3:33 ` Obsoleted large bucket layout message once Coly Li

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=de528a90-6d3d-d324-ca22-73b635d5153f@cssbook.de \
    --to=c.runge@cssbook.de \
    --cc=linux-bcache@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).