- 1. xfs_repair on large fs: out of memory (score: 1)
- Author: "Ed L. Cashin" <ecashin@xxxxxxxxxx>
- Date: Wed, 17 May 2006 15:36:06 -0400
- Hi. In trying to run xfs_repair on a 13 TB filesystem, the program either runs out of memory ... host02:~# sysctl vm.overcommit_memory=2 vm.overcommit_memory = 2 host02:~# /opt/xfsprogs-2.7.11/sbin/x
- /archives/xfs/2006-05/msg00025.html (8,545 bytes)
- 2. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: Chris Wedgwood <cw@xxxxxxxx>
- Date: Wed, 17 May 2006 15:23:53 -0700
- [...] I forget the ratio/math for working this out (Steve Lord posted a summary of this some time ago) but 1GB isn't nearly enough to run xfs_repair on a moderately sized filesystem.
- /archives/xfs/2006-05/msg00026.html (7,879 bytes)
- 3. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: Chris Wedgwood <cw@xxxxxxxx>
- Date: Wed, 17 May 2006 15:33:41 -0700
- Of course it should still not fail the way you describe with overcommit. Is there an rlimit set that is biting you?
- /archives/xfs/2006-05/msg00027.html (7,948 bytes)
- 4. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: Chris Wedgwood <cw@xxxxxxxx>
- Date: Wed, 17 May 2006 16:40:07 -0700
- I'm not sure this helps, but you got me thinking and I made a quick test here. On a near empty 13TB fs (I got bored after copying about 20GB of data on to it) I see xfs_repair here use a little under
- /archives/xfs/2006-05/msg00029.html (8,061 bytes)
- 5. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: pg_xfs@xxxxxxxxxxxxxxxxxx (Peter Grandi)
- Date: Thu, 18 May 2006 00:54:47 +0100
- I think I am not alone on this list in being very curious how long does the check+repair take. Also, how do you backup that filesystem? How long does it take? The summary is this one probably, from
- /archives/xfs/2006-05/msg00034.html (8,893 bytes)
- 6. xfs_repair on large fs: out of memory (score: 1)
- Author: "Ed L. Cashin" <ecashin@xxxxxxxxxx>
- Date: Wed, 17 May 2006 15:36:06 -0400
- Hi. In trying to run xfs_repair on a 13 TB filesystem, the program either runs out of memory ... host02:~# sysctl vm.overcommit_memory=2 vm.overcommit_memory = 2 host02:~# /opt/xfsprogs-2.7.11/sbin/x
- /archives/xfs/2006-05/msg00125.html (8,545 bytes)
- 7. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: Chris Wedgwood <cw@xxxxxxxx>
- Date: Wed, 17 May 2006 15:23:53 -0700
- [...] I forget the ratio/math for working this out (Steve Lord posted a summary of this some time ago) but 1GB isn't nearly enough to run xfs_repair on a moderately sized filesystem.
- /archives/xfs/2006-05/msg00126.html (7,879 bytes)
- 8. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: Chris Wedgwood <cw@xxxxxxxx>
- Date: Wed, 17 May 2006 15:33:41 -0700
- Of course it should still not fail the way you describe with overcommit. Is there an rlimit set that is biting you?
- /archives/xfs/2006-05/msg00127.html (7,948 bytes)
- 9. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: Chris Wedgwood <cw@xxxxxxxx>
- Date: Wed, 17 May 2006 16:40:07 -0700
- I'm not sure this helps, but you got me thinking and I made a quick test here. On a near empty 13TB fs (I got bored after copying about 20GB of data on to it) I see xfs_repair here use a little under
- /archives/xfs/2006-05/msg00129.html (8,061 bytes)
- 10. Re: xfs_repair on large fs: out of memory (score: 1)
- Author: pg_xfs@xxxxxxxxxxxxxxxxxx (Peter Grandi)
- Date: Thu, 18 May 2006 00:54:47 +0100
- I think I am not alone on this list in being very curious how long does the check+repair take. Also, how do you backup that filesystem? How long does it take? The summary is this one probably, from
- /archives/xfs/2006-05/msg00134.html (8,893 bytes)
This search system is powered by
Namazu