We have 2 linux machines here; the main one (RH 7.2) NFS mounts some
directories on the old one (RH 6.2), for nightly backups to CDRW. I had
to upgrade nfs-utils to 0.3.1 on the 6.2 machine to get that working.
(My wife *hates* to have her system changed, so a RH upgrade isn't an
option. Gotta keep the end user happy!)
But the 6.2 machine can't mount /home on the 7.2 machine (called posh):
$ mount /home/posh
mount: wrong fs type, bad option, bad superblock on posh:/home,
or too many mounted file systems
# tail -23 /var/log/messages
Jan 1 13:43:05 coo kernel: nfs: RPC call returned error 111
Jan 1 13:43:05 coo kernel: RPC: task of released request still queued!
Jan 1 13:43:05 coo kernel: RPC: (task is on xprt_pending)
Jan 1 13:43:05 coo kernel: nfs_read_super: get root fattr failed
Note, the 6.2 machine though gives this response to pmap_dump, which
looks good:
# /usr/sbin/pmap_dump
100000 2 tcp 111 portmapper
100000 2 udp 111 portmapper
100021 1 udp 1024 nlockmgr
100021 3 udp 1024 nlockmgr
100021 1 tcp 1024 nlockmgr
100021 3 tcp 1024 nlockmgr
100024 1 udp 1025 status
100024 1 tcp 1025 status
100011 1 udp 825 rquotad
100011 2 udp 825 rquotad
100003 2 udp 2049 nfs
100001 3 udp 866 rstatd
100001 2 udp 866 rstatd
100001 1 udp 866 rstatd
100005 1 udp 1027 mountd
100005 1 tcp 1028 mountd
100005 2 udp 1027 mountd
100005 2 tcp 1028 mountd
The 7.2 machine doesn't look so good (no mention of nfs):
# /usr/sbin/pmap_dump
100000 2 tcp 111 portmapper
100000 2 udp 111 portmapper
391002 2 tcp 32769 sgi_fam
100021 1 udp 32770 nlockmgr
100021 3 udp 32770 nlockmgr
100021 4 udp 32770 nlockmgr
100005 1 udp 32771 mountd
100005 1 tcp 32799 mountd
100005 2 udp 32771 mountd
100005 2 tcp 32799 mountd
100005 3 udp 32771 mountd
100005 3 tcp 32799 mountd
100024 1 udp 32779 status
100024 1 tcp 32981 status
/etc/rpc though does have the same nfs entries on both machines.
I notice there's no nfs daemon running on the 7.2 system. I'm running
at init level 3, so I just added a symlink from /etc/rc/d/rc3.d/S60nfs
to the init.d/nfs, but even doing an init.d/nfslock restart, no nfsd
process starts up, and the pmap_dump doesn't change. /var/log/messages
doesn't report anything odd:
Jan 1 14:48:18 posh rpc.statd[2547]: Caught signal 15, un-registering and exiting.
Jan 1 14:48:19 posh nfslock: rpc.statd shutdown succeeded
Jan 1 14:48:19 posh rpc.statd[2711]: Version 0.3.1 Starting
Jan 1 14:48:19 posh nfslock: rpc.statd startup succeeded
6.2 machine:
# ps ax | egrep "(rpc|nfs)"
390 ? SW 0:00 [rpciod]
399 ? S 0:00 rpc.statd
648 ? S 0:00 rpc.rquotad
666 ? SW 0:00 [nfsd]
667 ? SW 0:00 [nfsd]
668 ? SW 0:00 [nfsd]
669 ? SW 0:00 [nfsd]
670 ? SW 0:00 [nfsd]
671 ? SW 0:00 [nfsd]
672 ? SW 0:00 [nfsd]
673 ? SW 0:00 [nfsd]
687 ? S 0:00 rpc.rstatd
1053 ? S 0:00 rpc.mountd --no-nfs-version 3
1200 pts/3 S 0:00 egrep (rpc|nfs)
7.2 machine:
]# ps ax | egrep "(rpc|nfs)"
2141 ? SW 0:00 [rpciod]
2673 ? S 0:00 rpc.mountd
2711 ? S 0:00 rpc.statd
Incidentally, the --no-nfs-version 3 on the 6.2 machine is because
init.d/nfs explicitly adds that option for kernels earlier than 2.2.18,
and the 6.2 machine is running 2.2.14. But that doesn't explain why
the 7.2 machine doesn't seem to be running nfsd at all.
Any suggestions? How would you diagnose problems like this? I've been
doing google searches, which lead me to fix a couple of things (like,
an empty /etc/exports file on the 7.2 machine!), but finally I'm
stumped.
luke
--
SLUG - Sydney Linux User's Group - http://slug.org.au/
More Info: http://lists.slug.org.au/listinfo/slug
