Subject: kern/9653: Minor NFS3 problem between NetBSD and Digital UNIX
To: None <firstname.lastname@example.org>
From: Mark Davies <mark@MCS.VUW.AC.NZ>
Date: 03/21/2000 16:39:03
>Synopsis: Digital UNIX NFS servers report stale file handle problems from NetBSD clients
>Responsible: kern-bug-people (Kernel Bug People)
>Arrival-Date: Tue Mar 21 16:39:00 2000
>Originator: Mark Davies
Dept. of Comp. Sci., Victoria Uni. of Wellington, New Zealand.
System: NetBSD city-art.mcs.vuw.ac.nz 1.4S NetBSD 1.4S (MCS_WORKSTATION) #2: Thu Mar 16 09:07:41 NZDT 2000 email@example.com:/src/work/X/src/sys/arch/i386/compile/MCS_WORKSTATION i386
We have a number of NetBSD-current workstations that mount users home
directories off Alphas running Digital UNIX 4.0E. The syslog's on the
Alpha boxes quite often have entries like the following:
Mar 22 12:18:54 circa.mcs.vuw.ac.nz vmunix: RFS3_COMMIT, client address = 18.104.22.168, errno 22
Mar 22 12:20:32 mid-city.mcs.vuw.ac.nz vmunix: NFS server: stale file handle fs(8,1024) file 28572 gen 953004928
Mar 22 12:20:32 mid-city.mcs.vuw.ac.nz vmunix: RFS3_COMMIT, client address = 22.214.171.124, errno 22
Mar 22 12:22:06 majestic.mcs.vuw.ac.nz vmunix: NFS server: stale file handle fs(8,2055) file 196679 gen 952401980
Mar 22 12:22:06 majestic.mcs.vuw.ac.nz vmunix: RFS3_COMMIT, client address = 126.96.36.199, errno 22
Mar 22 12:26:02 greta-pt.mcs.vuw.ac.nz vmunix: NFS server: stale file handle fs(8,1024) file 119764 gen 953682960
Mar 22 12:26:02 greta-pt.mcs.vuw.ac.nz vmunix: RFS3_COMMIT, client address = 188.8.131.52, errno 22
Mar 22 12:26:24 bats.mcs.vuw.ac.nz vmunix: NFS server: stale file handle fs(8,36869) file 35790 gen 953891031
Mar 22 12:26:24 bats.mcs.vuw.ac.nz vmunix: RFS3_GETATTR, client address = 184.108.40.206, errno 22
The client address is always a NetBSD box.
I haven't noticed any actual problems accessing files related to this
but these messages are filling up my logs. Is this a bug with
NetBSD's NFS or with the Alphas? How could I track it down?
I'm running a current from mid Feb but the same behaviour was there
with previous currents I've run.