Re: [PATCH] Cleancache: shim to Xen Transcendent Memory

2010-07-12 Thread Konrad Rzeszutek Wilk
On Thu, Jul 08, 2010 at 09:42:08AM -0700, Dan Magenheimer wrote:
 Signed-off-by: Dan Magenheimer dan.magenhei...@oracle.com

One nitpick:
..
 +
 +int tmem_enabled;
 +
 +static int __init enable_tmem(char *s)
 +{
 + tmem_enabled = 1;
 + return 1;
 +}
 +
 +__setup(tmem, enable_tmem);

Perhaps 'tmem_setup' is more appropiate as it might be that this
function in the future would be only used to disable tmem, not actually
enable it?

Otherwise, it has been reviewed by me.
--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: slow deletion of files

2010-07-12 Thread Clemens Eisserer
Forgot to mention, I filed already a report for this problem:

https://bugzilla.kernel.org/show_bug.cgi?id=16117

However, like the other btrfs bug I filed, it was never looked at - so
I decided it was a waste of time to file bugs against it.

- Clemens
--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: slow deletion of files

2010-07-12 Thread Tracy Reed
On Mon, Jul 12, 2010 at 10:30:20PM +0200, Clemens Eisserer spake thusly:
 Another reason I moved away was btrfs corrupted, and btrfsck is still
 not able to repair it.
 I really like btrfs but in my opinion it has still a long road to go
 and declaring it stable in 2.6.35 is quite optimistic at best.

How many of reiserfs' problems were due to bugs in reiserfs vs due to
buggy PC memory which is rarely ECC? These fancy new filesystems hold
a lot of datastructures in memory compared to older filesystems which
would seem to increase the chances that they could be broken by bad
RAM. I am concerned that a flipped bit in memory somewhere could be
written out to disk and hose the filesystem. I know ZFS implements a
lot of checksums to prevent this sort of thing but it also tends to
run on nicer hardware with ECC. I never had corruption problems with
reiserfs even while running it on many terabytes of disk. I know
plenty of people who constantly lost data to it. I can't explain the
difference other than hardware.

-- 
Tracy Reed
http://tracyreed.org


pgpyK8k8VpKSy.pgp
Description: PGP signature


Re: Disk space usage displayed incorrectly?

2010-07-12 Thread Chris Mason
On Sat, Jul 10, 2010 at 07:22:26AM +0530, jck wrote:
 Hello,
 when i check diskspace usage of a btrfs partition using df i get the
 wrong free space, this is expected i think.

We do try to take the raid levels into account during df, but only for
the space that is already allocated.

 However even when i use 'btrfs filesystem df' I get wrong freespace:
 Data: total=123.58GB, used=87.31GB
 Metadata: total=61.00GB, used=396.29MB
 System: total=32.00MB, used=16.00KB
 
 Does this mean that after the 123 GB of 'Data' fills up I wont be able
 to add more stuff to the partition?
 I'm using btrfs-progs-git (jul 10)

This does reflect the space that is allocated, but the drive itself may
have more free space.  btrfs filesystem show will give you a few more
details.

Basically btrfs allocates from the drive in chunks of about 1GB and then
puts the file data or metadata into those chunks.  So if you do a mkfs
and then create one file, you might have 1TB of space on the drive that
hasn't been allocated yet.

-chris

--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: Possible corruption following a crash

2010-07-12 Thread Chris Mason
On Sun, Jul 11, 2010 at 12:39:58AM +0300, nokia wrote:
  http://pastebin.com/5r07k9Tn
 
 
 Hi everyone,
 
 A power outage occurring after successfully converting
 ext3-btrfs/mounting an external drive now freezes at least on .32
 kernels (Lucid and F12) with the traceback shown in the pastebin.
 Following cjb's instructions in #btrfs I'm reporting it here.

Thanks, could you please include the error messages during mount?

-chris
--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: Disk space usage displayed incorrectly?

2010-07-12 Thread jck
I tested it and the partition gets filled at 123.58 GB, the 60GB is
getting wasted. is there anything i can do about this?

On Tue, Jul 13, 2010 at 6:46 AM, Chris Mason chris.ma...@oracle.com wrote:
 On Sat, Jul 10, 2010 at 07:22:26AM +0530, jck wrote:
 Hello,
 when i check diskspace usage of a btrfs partition using df i get the
 wrong free space, this is expected i think.

 We do try to take the raid levels into account during df, but only for
 the space that is already allocated.

 However even when i use 'btrfs filesystem df' I get wrong freespace:
 Data: total=123.58GB, used=87.31GB
 Metadata: total=61.00GB, used=396.29MB
 System: total=32.00MB, used=16.00KB

 Does this mean that after the 123 GB of 'Data' fills up I wont be able
 to add more stuff to the partition?
 I'm using btrfs-progs-git (jul 10)

 This does reflect the space that is allocated, but the drive itself may
 have more free space.  btrfs filesystem show will give you a few more
 details.

 Basically btrfs allocates from the drive in chunks of about 1GB and then
 puts the file data or metadata into those chunks.  So if you do a mkfs
 and then create one file, you might have 1TB of space on the drive that
 hasn't been allocated yet.

 -chris


--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: Disk space usage displayed incorrectly?

2010-07-12 Thread jck
I ran btrfs filesystem balance but now it has become like this:


Metadata: total=5.00GB, used=507.25MB
Data: total=113.58GB, used=89.51GB
System: total=32.00MB, used=24.00KB

On Tue, Jul 13, 2010 at 7:23 AM, jck j...@archlinux.us wrote:
 I tested it and the partition gets filled at 123.58 GB, the 60GB is
 getting wasted. is there anything i can do about this?

 On Tue, Jul 13, 2010 at 6:46 AM, Chris Mason chris.ma...@oracle.com wrote:
 On Sat, Jul 10, 2010 at 07:22:26AM +0530, jck wrote:
 Hello,
 when i check diskspace usage of a btrfs partition using df i get the
 wrong free space, this is expected i think.

 We do try to take the raid levels into account during df, but only for
 the space that is already allocated.

 However even when i use 'btrfs filesystem df' I get wrong freespace:
 Data: total=123.58GB, used=87.31GB
 Metadata: total=61.00GB, used=396.29MB
 System: total=32.00MB, used=16.00KB

 Does this mean that after the 123 GB of 'Data' fills up I wont be able
 to add more stuff to the partition?
 I'm using btrfs-progs-git (jul 10)

 This does reflect the space that is allocated, but the drive itself may
 have more free space.  btrfs filesystem show will give you a few more
 details.

 Basically btrfs allocates from the drive in chunks of about 1GB and then
 puts the file data or metadata into those chunks.  So if you do a mkfs
 and then create one file, you might have 1TB of space on the drive that
 hasn't been allocated yet.

 -chris



--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html


Re: Disk space usage displayed incorrectly?

2010-07-12 Thread jck
I finally understood whats going on, thank you.

On Tue, Jul 13, 2010 at 9:32 AM, jck j...@archlinux.us wrote:
 I ran btrfs filesystem balance but now it has become like this:


 Metadata: total=5.00GB, used=507.25MB
 Data: total=113.58GB, used=89.51GB
 System: total=32.00MB, used=24.00KB

 On Tue, Jul 13, 2010 at 7:23 AM, jck j...@archlinux.us wrote:
 I tested it and the partition gets filled at 123.58 GB, the 60GB is
 getting wasted. is there anything i can do about this?

 On Tue, Jul 13, 2010 at 6:46 AM, Chris Mason chris.ma...@oracle.com wrote:
 On Sat, Jul 10, 2010 at 07:22:26AM +0530, jck wrote:
 Hello,
 when i check diskspace usage of a btrfs partition using df i get the
 wrong free space, this is expected i think.

 We do try to take the raid levels into account during df, but only for
 the space that is already allocated.

 However even when i use 'btrfs filesystem df' I get wrong freespace:
 Data: total=123.58GB, used=87.31GB
 Metadata: total=61.00GB, used=396.29MB
 System: total=32.00MB, used=16.00KB

 Does this mean that after the 123 GB of 'Data' fills up I wont be able
 to add more stuff to the partition?
 I'm using btrfs-progs-git (jul 10)

 This does reflect the space that is allocated, but the drive itself may
 have more free space.  btrfs filesystem show will give you a few more
 details.

 Basically btrfs allocates from the drive in chunks of about 1GB and then
 puts the file data or metadata into those chunks.  So if you do a mkfs
 and then create one file, you might have 1TB of space on the drive that
 hasn't been allocated yet.

 -chris




--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html