Messages in this thread | | | Date | Thu, 3 Sep 1998 01:39:38 -0700 (PDT) | From | Simon Kirby <> | Subject | Re: NFS: Hash collision with the root inode ??? |
| |
On Wed, 2 Sep 1998, Matthias Urlichs wrote:
> I have a VERY strange NFS error here. > > Whenever I lstat() a certain file across NFS, the call succeeds, but all > accesses afterwards return EIO. The kernel logs: > > __nfs_fhget: inode 1234567 busy, i_count=2, i_nlink=18 > __nfs_fhget: inode 1234567 still busy, i_count=2 > __nfs_fhget: killing /// filehandle > > After this, the mount is dead; everything afterwards returns -EIO. > This happens only with this one specific file and is 100% reproducible.
Is your mount point turning into the file you are mentioning? If so, I think I know what's happening. I had the same problem over here...I wrote the NFS guys and they said that it is a problem with the hashed inode numbers -- in order to avoid the problem, you have to mount every partition that is under your mounted partition via NFS, not just the root. Of course, if there are no more filesystems mouned under the root of your NFS mount on the remote server, this won't help. I'm hoping you do have some, though. :)
Here's a script I wrote that, after mounting the server's root, will read /etc/fstab and do the appropriate mounts for everything else:
[sroot@alfie:/dock]# cat unix_mount #!/bin/tcsh -f mount -t nfs "${*}:/" "$*/live" || exit -1 echo " ${*}:/ mounted on $*/live/" foreach a (`cat "$*/live/etc/mtab" | egrep '^[^ ]+ /[^ ]+ ext2 ' | cut -f2 -d' '`) mount -t nfs "${*}:$a" "$*/live$a/" echo " ${*}:$a mounted on $*/live$a/" end Hope this helps. :)
Simon-
| Simon Kirby | Systems Administration | | mailto:sim@netnation.com | NetNation Communications | | http://www.netnation.com/ | Tech: (604) 684-6892 |
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html
| |