* "free space cache generation" ?
@ 2012-09-19 8:05 Swâmi Petaramesh
2012-09-19 16:01 ` Josef Bacik
0 siblings, 1 reply; 6+ messages in thread
From: Swâmi Petaramesh @ 2012-09-19 8:05 UTC (permalink / raw)
To: Linux Btrfs
Hi there,
I noticed my syslog is filling up with tons of:
btrfs: free space inode generation (0) did not match free space cache
generation (25326)
...any clue ?
TIA.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: "free space cache generation" ?
2012-09-19 8:05 "free space cache generation" ? Swâmi Petaramesh
@ 2012-09-19 16:01 ` Josef Bacik
2012-09-19 17:30 ` Swâmi Petaramesh
0 siblings, 1 reply; 6+ messages in thread
From: Josef Bacik @ 2012-09-19 16:01 UTC (permalink / raw)
To: Swâmi Petaramesh; +Cc: Linux Btrfs
On Wed, Sep 19, 2012 at 02:05:41AM -0600, Swâmi Petaramesh wrote:
> Hi there,
>
> I noticed my syslog is filling up with tons of:
>
> btrfs: free space inode generation (0) did not match free space cache
> generation (25326)
>
Yeah I need to just put those under a DEBUG ifdef, they aren't a big deal, just
letting you know that the on disk cache is invalid and it will have to put it
together ye olde fashioned way. Thanks,
Josef
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: "free space cache generation" ?
2012-09-19 16:01 ` Josef Bacik
@ 2012-09-19 17:30 ` Swâmi Petaramesh
2012-09-19 17:36 ` Josef Bacik
0 siblings, 1 reply; 6+ messages in thread
From: Swâmi Petaramesh @ 2012-09-19 17:30 UTC (permalink / raw)
To: Josef Bacik; +Cc: Linux Btrfs
Le 19/09/2012 18:01, Josef Bacik a écrit :
> Yeah I need to just put those under a DEBUG ifdef, they aren't a big
> deal, just letting you know that the on disk cache is invalid and it
> will have to put it together ye olde fashioned way.
...But is it supposed to, and how to fix it ?
--
Swâmi Petaramesh <swami@petaramesh.org> http://petaramesh.org PGP 9076E32E
Ne cherchez pas : Je ne suis pas sur Facebook.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: "free space cache generation" ?
2012-09-19 17:30 ` Swâmi Petaramesh
@ 2012-09-19 17:36 ` Josef Bacik
2012-09-19 17:54 ` Swâmi Petaramesh
0 siblings, 1 reply; 6+ messages in thread
From: Josef Bacik @ 2012-09-19 17:36 UTC (permalink / raw)
To: Swâmi Petaramesh; +Cc: Josef Bacik, Linux Btrfs
On Wed, Sep 19, 2012 at 11:30:39AM -0600, Swâmi Petaramesh wrote:
> Le 19/09/2012 18:01, Josef Bacik a écrit :
> > Yeah I need to just put those under a DEBUG ifdef, they aren't a big
> > deal, just letting you know that the on disk cache is invalid and it
> > will have to put it together ye olde fashioned way.
>
> ...But is it supposed to, and how to fix it ?
It can happen for all sorts of different reasons. Did you have a unclean
unmount at some point? Do you have a particularly full file system? I recently
increased the size of the free space cache, so if you are not on a kernel with
that patch you are more likely to run out of space while writing out free space
cache and run into this problem. Thanks,
Josef
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: "free space cache generation" ?
2012-09-19 17:36 ` Josef Bacik
@ 2012-09-19 17:54 ` Swâmi Petaramesh
2012-09-19 18:02 ` Josef Bacik
0 siblings, 1 reply; 6+ messages in thread
From: Swâmi Petaramesh @ 2012-09-19 17:54 UTC (permalink / raw)
To: Josef Bacik; +Cc: Linux Btrfs
Le 19/09/2012 19:36, Josef Bacik a écrit :
> It can happen for all sorts of different reasons. Did you have a
> unclean unmount at some point?
Not that I'm aware of...
> Do you have a particularly full file system?
~1 TB FS, 23% full...
Only, I have several Linux distros installed on the same BTRFS - in
different subvols - so I sometimes mount it with a 3.5.x kernel, and
more rarely - but sometimes - with a 3.2.x...
=> Also I have to mention that I have lost a complete BTRFS FS, that was
a /boot with snapshots. At some point it started to fail mounting, the
kernel complaining that some trees (corresponding to removed snapshots
as far as I could tell) were b0rked, and btrfsck couldn't help...
So I lost the FS - and lost quite a bit of confidence into BTRFS the
same day... :-\
--
Swâmi Petaramesh <swami@petaramesh.org> http://petaramesh.org PGP 9076E32E
Ne cherchez pas : Je ne suis pas sur Facebook.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: "free space cache generation" ?
2012-09-19 17:54 ` Swâmi Petaramesh
@ 2012-09-19 18:02 ` Josef Bacik
0 siblings, 0 replies; 6+ messages in thread
From: Josef Bacik @ 2012-09-19 18:02 UTC (permalink / raw)
To: Swâmi Petaramesh; +Cc: Josef Bacik, Linux Btrfs
On Wed, Sep 19, 2012 at 11:54:38AM -0600, Swâmi Petaramesh wrote:
> Le 19/09/2012 19:36, Josef Bacik a écrit :
> > It can happen for all sorts of different reasons. Did you have a
> > unclean unmount at some point?
>
> Not that I'm aware of...
>
> > Do you have a particularly full file system?
>
> ~1 TB FS, 23% full...
>
> Only, I have several Linux distros installed on the same BTRFS - in
> different subvols - so I sometimes mount it with a 3.5.x kernel, and
> more rarely - but sometimes - with a 3.2.x...
>
Well I think I changed the space cache format between those releases which would
cause them all to be discarded and remade.
> => Also I have to mention that I have lost a complete BTRFS FS, that was
> a /boot with snapshots. At some point it started to fail mounting, the
> kernel complaining that some trees (corresponding to removed snapshots
> as far as I could tell) were b0rked, and btrfsck couldn't help...
>
> So I lost the FS - and lost quite a bit of confidence into BTRFS the
> same day... :-\
>
Yes well, welcome to using an experimental file system.
Josef
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2012-09-19 18:02 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-09-19 8:05 "free space cache generation" ? Swâmi Petaramesh
2012-09-19 16:01 ` Josef Bacik
2012-09-19 17:30 ` Swâmi Petaramesh
2012-09-19 17:36 ` Josef Bacik
2012-09-19 17:54 ` Swâmi Petaramesh
2012-09-19 18:02 ` Josef Bacik
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).