inode number mismatch message

inode number mismatch message

[Kent, Ian I.]

I recently got a rash of messages like these in the logs on our Linux
clients:

Mar 24 04:02:00 csl024 syslogd 1.4-0: restart.
Mar 24 04:02:00 csl024 syslogd 1.4-0: restart.
Mar 24 23:50:13 csl024 kernel: nfs: server perseus is not responding
Mar 24 23:50:28 csl024 kernel: nfs: server perseus OK
Mar 24 23:51:43 csl024 kernel: nfs_refresh_inode: inode number mismatch
Mar 24 23:51:43 csl024 kernel: expected (0x2813c6d/0x4), got
(0xa000013c6d/0x4)
Mar 24 23:51:43 csl024 kernel: nfs_refresh_inode: inode number mismatch
Mar 24 23:51:43 csl024 kernel: expected (0x2813c69/0x4), got
(0xa000013c69/0x4)
Mar 24 23:52:58 csl024 kernel: nfs_refresh_inode: inode number mismatch
Mar 24 23:52:58 csl024 kernel: expected (0x2813c6d/0x4), got
(0xa000013c6d/0x4)
Mar 24 23:52:58 csl024 kernel: nfs_refresh_inode: inode number mismatch
Mar 24 23:52:58 csl024 kernel: expected (0x2813c69/0x4), got
(0xa000013c69/0x4)

They appeared after the server (Solaris) to which the Linux (2.4.17 kernel
with Tronds' patches) was upgraded from 2.6 to 2.8 while there were active
mounts.

I know it's a big ask to expect the client to be OK though such an activity,
but the comment is 'the Solaris clients were OK after the upgrade'.

I would appreciate an feedback on whether this is something that should
fixable or not.

Regards 
Ian Kent

_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs

Re: inode number mismatch message

[James Pearson]

A previous message from Trond had a workaround:

http://marc.theaimsgroup.com/?l=linux-nfs&m=101316526524213&w=2

I believe this change is in Trond's 2.4.18 patches ...

James Pearson


"Kent, Ian I." wrote:
> 
> I recently got a rash of messages like these in the logs on our Linux
> clients:
> 
> Mar 24 04:02:00 csl024 syslogd 1.4-0: restart.
> Mar 24 04:02:00 csl024 syslogd 1.4-0: restart.
> Mar 24 23:50:13 csl024 kernel: nfs: server perseus is not responding
> Mar 24 23:50:28 csl024 kernel: nfs: server perseus OK
> Mar 24 23:51:43 csl024 kernel: nfs_refresh_inode: inode number mismatch
> Mar 24 23:51:43 csl024 kernel: expected (0x2813c6d/0x4), got
> (0xa000013c6d/0x4)
> Mar 24 23:51:43 csl024 kernel: nfs_refresh_inode: inode number mismatch
> Mar 24 23:51:43 csl024 kernel: expected (0x2813c69/0x4), got
> (0xa000013c69/0x4)
> Mar 24 23:52:58 csl024 kernel: nfs_refresh_inode: inode number mismatch
> Mar 24 23:52:58 csl024 kernel: expected (0x2813c6d/0x4), got
> (0xa000013c6d/0x4)
> Mar 24 23:52:58 csl024 kernel: nfs_refresh_inode: inode number mismatch
> Mar 24 23:52:58 csl024 kernel: expected (0x2813c69/0x4), got
> (0xa000013c69/0x4)
> 
> They appeared after the server (Solaris) to which the Linux (2.4.17 kernel
> with Tronds' patches) was upgraded from 2.6 to 2.8 while there were active
> mounts.
> 
> I know it's a big ask to expect the client to be OK though such an activity,
> but the comment is 'the Solaris clients were OK after the upgrade'.
> 
> I would appreciate an feedback on whether this is something that should
> fixable or not.
> 
> Regards
> Ian Kent
> 
> _______________________________________________
> NFS maillist  -  NFS@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/nfs

_______________________________________________
NFS maillist  -  NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs