preprocess_seqid_op: bad seqid (expected 2, got 50)

[Bryce Harrington]

Hi Bruce,

I'm working my way through the testruns that ran while I was gone last
month.  Most of the errors are either known things already reported,
but I noticed this one that looks interesting.

 http://crucible.osdl.org/runs/3839/sysinfo/nfs02.console
 This is nfs02.nfsv4.pdx.osdl.net (Linux i686
2.6.21-rc4-g992f541-nfs-server-stable) 21:33:35

 nfs02 login: nfsd: last server has exited
 nfsd: unexporting all filesystems
 NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory
 NFSD: starting 90-second grace period
 NFSD: preprocess_seqid_op: bad seqid (expected 2, got 50)
 NFSD: preprocess_seqid_op: magic stateid!
 NFSD: preprocess_seqid_op: old stateid!
 nfs4_cb: server 0.0.0.0 not responding, timed out
 nfs4_cb: server 0.0.0.0 not responding, timed out
 nfs4_cb: server 0.0.0.0 not responding, timed out
 nfs4_cb: server 0.0.0.0 not responding, timed out



Googling around I found one other person that also ran into this:

 http://www.mail-archive.com/scientific-linux-users@listserv.fnal.gov/msg00098.html

Unfortunately, since the errors aren't timestamped I can't be
absolutely positive which test was running when the above occurred,
however cthon appears to have gotten stuck while running against krb5,
so my best guess is it occurred with that run:

 http://crucible.osdl.org/runs/3839/test_output/cthon.krb5.log

 test freesp and file size
 fcntl(...F_FREESP...) not available on this platform.

 write/read 30 MB file

In bugzilla, I see I've run into this message previously, although the
symptoms in that case were a bit different, so I'm not sure it's
necessarily the same bug:

 http://bugzilla.linux-nfs.org/show_bug.cgi?id=112

Does this issue look like one worth deeper investigation?  Any
thoughts that might help me narrow in on it better?

Bryce