BTRFS cache problem
Hi,
My BTRFS is mounted with space_cache,inode_cache , and I now get at boot time the following message :
“btrfs: free space inode generation (0) did not match free space cache generation (189086)”
Starting once with "clear_cache" doesn't seem to be able to fix the issue... Any clue ?
TIA.
Signature · 2 lines
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.
Hi Swami,
which kernel version are you running?
Regards,
Bart
Quoted 21 lines · Swâmi Petaramesh
Hi,
My BTRFS is mounted with space_cache,inode_cache , and I now get at boot time the following message :
“btrfs: free space inode generation (0) did not match free space cache generation (189086)”
Starting once with "
clear_cache" doesn't seem to be able to fix the issue... Any clue ?TIA.
--
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 1 line · Bart Noordervliet
which kernel version are you running?
Hi Bart,
# uname -srv Linux 3.5.0-18-generic #29-Ubuntu SMP Fri Oct 19 10:26:51 UTC 2012
Signature · 2 lines
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.
A message that is not in the archive (sent off-list or not yet received): CAGy7UtgrdGpC5uav0m0yX1oJ47OR6WsX0vOtKu4p5umS4hZUtw@mail.gmail.com
Quoted 5 lines · Bart Noordervliet
Quoted 2 lines
# uname -srv Linux 3.5.0-18-generic #29-Ubuntu SMP Fri Oct 19 10:26:51 UTC 2012Could you first try with either 3.6 or a recent 3.7-rc build please? You can easily add them to Ubuntu by using their kernel ppa:
http://kernel.ubuntu.com/~kernel-ppa/mainline/
I previously had bad times running "future" mainline kernels when returning to previous "usual" kernels, as BTRFS didn't seem to like this back-and-forth kernel play...
I'm not going to take risks with this machine, which is my main laptop which I badly need to be in working order tomorrow morning (and #me having slept at night... ;-)
Signature · 2 lines
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.
Quoted 3 lines · Swâmi Petaramesh
I'm not going to take risks with this machine, which is my main laptop which I badly need to be in working order tomorrow morning (and #me having slept at night... ;-)
Without running different kernels, I don't think we can work with you to figure out the cause of the problem, unless this is a known problem (it isn't to me). You didn't mention whether the filesystem actually mounts properly or not, but if it does then I don't think you'll have problems other than perhaps diminished performance.
May I also suggest that you maybe shouldn't be running an experimental filesystem on a laptop that you can't go without?
Quoted 2 lines · Bart Noordervliet
May I also suggest that you maybe shouldn't be running an experimental filesystem on a laptop that you can't go without?
I might well be retired before BTRFS is no more "experimental", but I need its features now...
Otherwise I could go with “ZFS on Linux” that has proven fairly excellent and much faster and more stable than BTRFS on the machine on which I've been using it for ~2 years...
I chosed to go for BTRFS on this machine because it's officially part of the Linux kernel and I would expect it to improve (quickly) over time...
But for now I have to admit that I regret my choice performance wise (which is terrible...) and stability wise (Already 2 reinstalls with complete format, if there's a next one, the next one will be ZFS...)
Kind regards.
Signature · 2 lines
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.
Quoted 10 lines · Swâmi Petaramesh
Hi,
My BTRFS is mounted with space_cache,inode_cache , and I now get at boot time the following message :
“btrfs: free space inode generation (0) did not match free space cache generation (189086)”
Starting once with "
clear_cache" doesn't seem to be able to fix the issue... Any clue ?
You only need to mount with space_cache once, and after that it's
"sticky". What's probably happening is that you're mounting with space_cache, which rebuilds the cache. Then, if you reboot before it's finished rebuilding the cache, the cache isn't complete, and (I think) it causes the above message -- which is harmless. The same applies to the inode cache options.
Try mounting with clear_cache *once*, then remounting with
space_cache,inode_cache *once*, and waiting for it to finish rebuilding the cache. Finally, remove all mention of the caches from your fstab and kernel command line before you next mount the FS.
Hugo.
Signature · 4 lines
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
PGP key: 515C238D from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
--- Great oxymorons of the world, no. 8: The Latest ---
In Proven Technology Quoted 10 lines · Hugo Mills
You only need to mount with
space_cacheonce, and after that it's "sticky". What's probably happening is that you're mounting withspace_cache, which rebuilds the cache. Then, if you reboot before it's finished rebuilding the cache, the cache isn't complete, and (I think) it causes the above message -- which is harmless. The same applies to the inode cache options. Try mounting withclear_cache*once*, then remounting with space_cache,inode_cache *once*, and waiting for it to finish rebuilding the cache. Finally, remove all mention of the caches from your fstab and kernel command line before you next mount the FS. Hugo.
Thanks very much Hugo, I'll give this a try. Should I assume that the cache clearing / building is complete as soon as the disk is idle ?
Signature · 2 lines
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.
Quoted 12 lines · Swâmi Petaramesh
Quoted 10 lines · Hugo Mills
You only need to mount with
space_cacheonce, and after that it's "sticky". What's probably happening is that you're mounting withspace_cache, which rebuilds the cache. Then, if you reboot before it's finished rebuilding the cache, the cache isn't complete, and (I think) it causes the above message -- which is harmless. The same applies to the inode cache options. Try mounting withclear_cache*once*, then remounting with space_cache,inode_cache *once*, and waiting for it to finish rebuilding the cache. Finally, remove all mention of the caches from your fstab and kernel command line before you next mount the FS. Hugo.
Quoted 2 lines
Thanks very much Hugo, I'll give this a try. Should I assume that the cache clearing / building is complete as soon as the disk is idle ?
I'm not sure. Probably. There is probably (I haven't checked
recently) a kernel process which does the rebuilding.
Hugo.
Signature · 4 lines
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
PGP key: 515C238D from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
--- Great oxymorons of the world, no. 8: The Latest ---
In Proven Technology Quoted 4 lines · Hugo Mills
Try mounting with
clear_cache*once*, then remounting with space_cache,inode_cache *once*, and waiting for it to finish rebuilding the cache. Finally, remove all mention of the caches from your fstab and kernel command line before you next mount the FS.
Unfortunately, same issue after applying your suggestions...
[ 752.943532] btrfs: free space inode generation (0) did not match free
space cache generation (203710)
Signature · 2 lines
Swâmi Petaramesh <[email protected]> http://petaramesh.org PGP 9076E32E Ne cherchez pas : Je ne suis pas sur Facebook.