Re: NFS stale mount after chroot...

Daniel Freedman (freedman@ccmr.cornell.edu)
Sun, 9 Dec 2001 21:18:34 -0500


On Sun, Dec 09, 2001, Daniel Freedman wrote:
>
> Hi,
>
> It seems like I can generate reproducible stale NFS mounts by mounting
> a partition, chroot'ing into that mount, immediately exiting the
> chroot, and then finding myself unable to unmount the NFS partition.
> I'm pretty sure I've confirmed that nothing is using the partition
> (both with fuser and lsof) and even tried to force umount the
> partition (which seems like it should definitely umount it, rather
> than returning with the same "device is busy" errors), to no avail.
> The only method which I've used that seems to be able to get rid of
> this NFS mount, is to reboot the NFS client, and clearly that's not a
> good one at all. If I'm missing something obvious here, my apologies
> in advance. Also, if there's any further information I can provide,
> I'd be happy to help. The dump of my procedure follows this message.
>
> Thanks again and take care,
> Daniel

Arggg... (Continuing above, appending to previous email's client
output, and sorry for leaving off some info...)

Here's my general system info, as well as a counterpoint successful
attempt (to hopefully narrow the search area) at unmounting the same
NFS fs, the only difference is that I didn't do a chroot into that
mount in the interim. Hope this helps as well.

Take care,
Daniel

--------
NFS Client:
________

feynman:/var/space/freedman# uname -a
Linux feynman 2.4.16 #1 SMP Wed Nov 28 12:48:20 EST 2001 i686 unknown
feynman:/var/space/freedman# cat /boot/config-2.4.16 |grep NFS
CONFIG_NFS_FS=m
CONFIG_NFS_V3=y
# CONFIG_ROOT_NFS is not set
CONFIG_NFSD=m
CONFIG_NFSD_V3=y
CONFIG_NCPFS_NFS_NS=y
feynman:/var/space/freedman# cat /proc/sys/kernel/tainted
0
feynman:/var/space/freedman# lsmod
Module Size Used by Not tainted
r128 91608 1
emu10k1 58880 0 (unused)
sound 56940 0 [emu10k1]
soundcore 4036 7 [emu10k1 sound]
ac97_codec 9760 0 [emu10k1]
nfs 73692 1
lockd 48704 1 [nfs]
sunrpc 65108 1 [nfs lockd]
eepro100 17584 1
rtc 6296 0 (autoclean)
unix 16004 19 (autoclean)
ide-disk 6880 8 (autoclean)
ide-probe-mod 8096 0 (autoclean)
ide-mod 132876 8 (autoclean) [ide-disk ide-probe-mod]
ext2 32960 5 (autoclean)
feynman:/var/space/freedman# mount -t nfs newton:/var/tftpboot-NFS/ ./node1/
feynman:/var/space/freedman# cat /proc/mounts
/dev/root.old /initrd cramfs rw 0 0
/dev/root / ext2 rw 0 0
proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/dev/hda5 /boot ext2 rw 0 0
/dev/hda7 /usr ext2 rw 0 0
/dev/hda8 /var ext2 rw 0 0
/dev/hda12 /var/space ext2 rw 0 0
newton:/home /home nfs rw,v3,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=newton 0 0
newton:/var/tftpboot-NFS/ /var/space/freedman/node1 nfs rw,v3,rsize=8192,wsize=8192,hard,udp,lock,addr=newton 0 0
feynman:/var/space/freedman# cd node1/
feynman:/var/space/freedman/node1# ls
bin boot cdrom dev etc floppy home initrd lib mnt opt proc root sbin tmp usr var vmlinuz
feynman:/var/space/freedman/node1# id
uid=0(root) gid=0(root) groups=0(root)
feynman:/var/space/freedman/node1# cd ..
feynman:/var/space/freedman# umount /var/space/freedman/node1/
feynman:/var/space/freedman# cat /proc/mounts
/dev/root.old /initrd cramfs rw 0 0
/dev/root / ext2 rw 0 0
proc /proc proc rw 0 0
devpts /dev/pts devpts rw 0 0
/dev/hda5 /boot ext2 rw 0 0
/dev/hda7 /usr ext2 rw 0 0
/dev/hda8 /var ext2 rw 0 0
/dev/hda12 /var/space ext2 rw 0 0
newton:/home /home nfs rw,v3,rsize=8192,wsize=8192,hard,intr,udp,lock,addr=newton 0 0
feynman:/var/space/freedman#

-- 
Daniel A. Freedman
Laboratory for Atomic and Solid State Physics
Department of Physics
Cornell University
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/