* btrfs: free space inode generation (0) did not match free space cache generation
@ 2011-01-28 9:59 Tomasz Chmielewski
2011-01-28 13:45 ` Josef Bacik
0 siblings, 1 reply; 4+ messages in thread
From: Tomasz Chmielewski @ 2011-01-28 9:59 UTC (permalink / raw)
To: linux-btrfs
Not sure if the:
btrfs: free space inode generation (0) did not match free space cache generation
is supposed to be fixed in 2.6.38-rc2?
I'm getting such entries in dmesg:
[12638.713154] device fsid 1142843480ad2d13-4bdc742fd9b1f7b0 devid 1 transid 3501 /dev/sdb4
[12638.713528] btrfs: force zlib compression
[12638.959864] btrfs: disk space caching is enabled
[12679.397549] btrfs: free space inode generation (0) did not match free space cache generation (3491) for block group 10766778368
[12853.160162] btrfs: free space inode generation (0) did not match free space cache generation (3467) for block group 282960330752
[12853.160176] btrfs: free space inode generation (0) did not match free space cache generation (3470) for block group 285107814400
[12853.160187] btrfs: free space inode generation (0) did not match free space cache generation (3478) for block group 290476523520
[12853.160198] btrfs: free space inode generation (0) did not match free space cache generation (3480) for block group 292624007168
[12853.237298] btrfs: free space inode generation (0) did not match free space cache generation (3485) for block group 296918974464
[12853.237316] btrfs: free space inode generation (0) did not match free space cache generation (3486) for block group 299066458112
[12853.237328] btrfs: free space inode generation (0) did not match free space cache generation (3487) for block group 300140199936
[12853.237339] btrfs: free space inode generation (0) did not match free space cache generation (3488) for block group 301213941760
[12858.468374] btrfs: free space inode generation (0) did not match free space cache generation (3430) for block group 223904530432
[12858.468392] btrfs: free space inode generation (0) did not match free space cache generation (3432) for block group 226052014080
[12858.468405] btrfs: free space inode generation (0) did not match free space cache generation (3434) for block group 228199497728
[12860.040819] btrfs: free space inode generation (0) did not match free space cache generation (3420) for block group 213167112192
[12860.040836] btrfs: free space inode generation (0) did not match free space cache generation (3422) for block group 215314595840
[12860.040852] btrfs: free space inode generation (0) did not match free space cache generation (3424) for block group 217462079488
[12860.040865] btrfs: free space inode generation (0) did not match free space cache generation (3426) for block group 219609563136
[12860.040877] btrfs: free space inode generation (0) did not match free space cache generation (3428) for block group 221757046784
I saw a similar discussion in December 2010, but I'm not sure if the fix was pushed yet or not?
The filesystem should have plenty of free space free:
Filesystem Size Used Avail Use% Mounted on
/dev/sdb4 336G 156G 179G 47% /srv
--
Tomasz Chmielewski
http://wpkg.org
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: btrfs: free space inode generation (0) did not match free space cache generation
2011-01-28 9:59 btrfs: free space inode generation (0) did not match free space cache generation Tomasz Chmielewski
@ 2011-01-28 13:45 ` Josef Bacik
2011-01-28 13:56 ` Tomasz Chmielewski
0 siblings, 1 reply; 4+ messages in thread
From: Josef Bacik @ 2011-01-28 13:45 UTC (permalink / raw)
To: Tomasz Chmielewski; +Cc: linux-btrfs
On Fri, Jan 28, 2011 at 10:59:56AM +0100, Tomasz Chmielewski wrote:
> Not sure if the:
>
> btrfs: free space inode generation (0) did not match free space cache generation
>
> is supposed to be fixed in 2.6.38-rc2?
>
It's not a problem, it just means that the cache is being discarded, which is
what is supposed to happen. You probably ran out of space or one of the other
corner cases where we can't write out the space cache. Thanks,
Josef
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: btrfs: free space inode generation (0) did not match free space cache generation
2011-01-28 13:45 ` Josef Bacik
@ 2011-01-28 13:56 ` Tomasz Chmielewski
2011-01-28 13:57 ` Josef Bacik
0 siblings, 1 reply; 4+ messages in thread
From: Tomasz Chmielewski @ 2011-01-28 13:56 UTC (permalink / raw)
To: Josef Bacik; +Cc: linux-btrfs
On 28.01.2011 14:45, Josef Bacik wrote:
> It's not a problem, it just means that the cache is being discarded, which is
> what is supposed to happen. You probably ran out of space or one of the other
> corner cases where we can't write out the space cache. Thanks,
FYI, there is plenty of free space (~150 GB).
It could be some corner case:
- filesystem is mounted with compress-force,
- one kvm process (the only writer to btrfs) writes to the btrfs
filesystem, via the virtual guest using a 220 GB qcow2 image file (lots
of writes) and a 7 GB qcow2 image file (very few writes).
--
Tomasz Chmielewski
http://wpkg.org
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: btrfs: free space inode generation (0) did not match free space cache generation
2011-01-28 13:56 ` Tomasz Chmielewski
@ 2011-01-28 13:57 ` Josef Bacik
0 siblings, 0 replies; 4+ messages in thread
From: Josef Bacik @ 2011-01-28 13:57 UTC (permalink / raw)
To: Tomasz Chmielewski; +Cc: Josef Bacik, linux-btrfs
On Fri, Jan 28, 2011 at 02:56:24PM +0100, Tomasz Chmielewski wrote:
> On 28.01.2011 14:45, Josef Bacik wrote:
>> It's not a problem, it just means that the cache is being discarded, which is
>> what is supposed to happen. You probably ran out of space or one of the other
>> corner cases where we can't write out the space cache. Thanks,
>
> FYI, there is plenty of free space (~150 GB).
>
> It could be some corner case:
>
> - filesystem is mounted with compress-force,
> - one kvm process (the only writer to btrfs) writes to the btrfs
> filesystem, via the virtual guest using a 220 GB qcow2 image file (lots
> of writes) and a 7 GB qcow2 image file (very few writes).
>
I'll try some tests with compress-force to make sure it's not getting overly
screwed, but this is just one of the drawbacks with the space cache stuff,
sometimes we just can't write it out without deadlocking, so we err on the side
of caution. Thanks,
Josef
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2011-01-28 13:57 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-01-28 9:59 btrfs: free space inode generation (0) did not match free space cache generation Tomasz Chmielewski
2011-01-28 13:45 ` Josef Bacik
2011-01-28 13:56 ` Tomasz Chmielewski
2011-01-28 13:57 ` Josef Bacik
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox