Messages in this thread | | | Date | Sun, 24 Apr 2011 22:51:17 -0700 (PDT) | From | Christian Kujau <> | Subject | Re: 2.6.39-rc4+: oom-killer busy killing tasks |
| |
On Mon, 25 Apr 2011 at 09:46, Dave Chinner wrote: > I'd say they are not being reclaimmmed because the VFS hasn't let go > of them yet. Can you also dump /proc/sys/fs/{dentry,inode}-state so > we can see if the VFS has released the inodes such that they can be > reclaimed by XFS?
Will do that.
> BTW, what are your mount options? If it is the problem I suspect it > is, then using noatime with stop it from occurring....
# grep xfs /proc/mounts /dev/mapper/wdc1 /mnt/backup xfs rw,nosuid,nodev,noexec,relatime,attr2,delaylog,noquota 0 0 /dev/mapper/wdc1 /data xfs ro,relatime,attr2,delaylog,noquota 0 0
The "/dev/mapper/wdc1" is a dm-crypt device on a ~1TB external disk (via FireWire). The 2nd line in the output above is a bind-mount from somewhere below "/mnt/backup".
I'm currently trying to reproduce the problem again and see if the {dentry,inode}-state output reveals any more information. After that, I'll try with noatime.
Thanks, Christian. -- BOFH excuse #115:
your keyboard's space bar is generating spurious keycodes.
| |