linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Leszek Dubiel <leszek@dubiel.pl>
To: linux-btrfs@vger.kernel.org
Subject: Re: enospc errors during balance — how to prevent out of space
Date: Mon, 29 Apr 2024 23:05:34 +0200	[thread overview]
Message-ID: <df6e9442-2144-462a-b339-1b385a1324ba@dubiel.pl> (raw)
In-Reply-To: <909ed27.f7893fdf.18f020ff6fc@tnonline.net>



Final report for "no space" and balancing.
I write it for maybe it will help someone.


My problem was identical to this one:

https://superuser.com/questions/1573030/cannot-repair-btrfs-partition-because-there-is-too-little-space-left



While mounting it went "read only" because of errors:

         errs: wr: 0, rd 0, flush 0, corrupt: 35967, gen 0

         "No space left failed to recover relocation"

         space_info: metadata ... is full



On "read only" you cannot add more devices, you canot finish balance.



Tried to repair system — no errors.



It was done on terminal, co I uploaded photos here:

https://postimg.cc/gallery/09dzGRG



Finally I have cleared disks and start to rebuild system from scratch.





Balancing script reworked to start balancing earlier and slower,
then go faster:



     findmnt --types btrfs --output SOURCE --nofsroot --noheadings | 
sort | uniq |
     while read dev; do

         # mount point
         mnt=$(findmnt --source "$dev" --output TARGET --first-only 
--noheadings)
         test -d "$mnt" || continue

         # lowest unallocated space from all disks in btrfs filesystem
         una=$(
             btrfs dev usage "$mnt" -g |
             sed -nr 's/.*Unallocated: +([0-9]+)\.[0-9]{2}GiB.*/\1/; T; p' |
             sort -n | head -n1
         )
         echo -n "$una" | tr -c "[[:print:]]" "#" | grep -Eq '^[0-9]+$' 
|| continue

         # if lots of unallocated space do nothing
         if test "$una" -ge 32; then
             : # do nothing

         # if going low, then optimize "dusage" slowly
         elif test "$una" -ge 16; then
             seq --format "-dusage=%g,limit=2" 0 20 100

         elif test "$una" -ge 8; then
             seq --format "-dusage=%g,limit=3" 0 10 100

         # if critically low, then balnace both "dusage" and "musage"
         else
             seq --format "-dusage=%g,limit=4" 0 10 100
             seq --format "-musage=%g,limit=1" 0 10 100
         fi |

         # balance for options above until any extents rellocated
         while read opt; do

             btrfs balance start "$opt" "$mnt" 2>&1 |
             grep -Eq "Done, had to relocate [1-9][0-9]* out of [0-9]+ 
chunks" &&

             # if relocated, then get out of two loops for next "$dev"
             break 2
         done
     done







      parent reply	other threads:[~2024-04-29 21:05 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-16  6:09 enospc errors during balance — how to prevent out of space Leszek Dubiel
2024-04-16  7:54 ` HAN Yuwei
2024-04-16 15:17   ` Leszek Dubiel
2024-04-16 17:00     ` Forza
2024-04-16 20:07       ` Leszek Dubiel
2024-04-21 19:09         ` Forza
2024-04-21 19:52           ` Leszek Dubiel
2024-04-27 12:03           ` Leszek Dubiel
2024-04-29 21:05           ` Leszek Dubiel [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=df6e9442-2144-462a-b339-1b385a1324ba@dubiel.pl \
    --to=leszek@dubiel.pl \
    --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).