Hello!
I have set up the most simple nfs-server configuration (I can imagine) and it shows funny behaviour:
Facts which seem relevant to me in this context:
0. I'm running on Fedora Core 1 (somewhat outdated, but well)
1. My nfs-server gets the IP via DHCP from a DSL Router via wireless lan. (192.168.1.11)
Its name "Vigor11" seems to come from the Router also because I configured it nowhere.
2. The /etc/exports looks like this:
"/misc *(rw)"
rpcinfo -p yields:
<snip>
100000 2 tcp 111 portmapper
100000 2 udp 111 portmapper
100024 1 udp 32768 status
100024 1 tcp 32770 status
391002 2 tcp 32771 sgi_fam
100011 1 udp 616 rquotad
100011 2 udp 616 rquotad
100011 1 tcp 619 rquotad
100011 2 tcp 619 rquotad
100003 2 udp 2049 nfs
100003 3 udp 2049 nfs
100003 2 tcp 2049 nfs
100003 3 tcp 2049 nfs
100021 1 udp 32807 nlockmgr
100021 3 udp 32807 nlockmgr
100021 4 udp 32807 nlockmgr
100021 1 tcp 32788 nlockmgr
100021 3 tcp 32788 nlockmgr
100021 4 tcp 32788 nlockmgr
100005 1 udp 634 mountd
100005 1 tcp 637 mountd
100005 2 udp 634 mountd
100005 2 tcp 637 mountd
100005 3 udp 634 mountd
100005 3 tcp 637 mountd
</snip>
4. exportfs yields:
"/misc <world>"
5. My /etc/hosts file looks like this (I currently work with IPs exclusively - if I get this
to work I will start using names):
"127.0.0.1 localhost.localdomain localhost"
Observations:
1. When I boot the server it takes several minutes to start the NFS-service (but it works!)
2. Directly after boot-up I get these lines regarding nfs in /var/log/messages:
<snip>
Mar 26 22:45:52 Vigor11 nfs: rpc.mountd shutdown failed
Mar 26 22:45:52 Vigor11 nfs: nfsd shutdown failed
Mar 26 22:45:52 Vigor11 nfs: rpc.rquotad shutdown failed
Mar 26 22:45:52 Vigor11 nfs: Shutting down NFS services: succeeded
Mar 26 22:45:52 Vigor11 exportfs[3122]: /etc/exports [1]: No 'sync' or 'async' option specified for export "*:/misc". Assuming default behaviour ('sync'). NOTE: this default has changed from previous versions
Mar 26 22:45:52 Vigor11 exportfs: exportfs: /etc/exports [1]: No 'sync' or 'async' option specified for export "*:/misc".
Mar 26 22:45:52 Vigor11 exportfs: Assuming default behaviour ('sync').
Mar 26 22:45:52 Vigor11 exportfs: NOTE: this default has changed from previous versions
Mar 26 22:47:12 Vigor11 kernel: Installing knfsd (copyright (C) 1996
okir@monad.swb.de).
Mar 26 22:47:12 Vigor11 nfs: Starting NFS services: succeeded
Mar 26 22:47:13 Vigor11 nfs: rpc.rquotad startup succeeded
Mar 26 22:47:13 Vigor11 nfs: rpc.nfsd startup succeeded
Mar 26 22:47:13 Vigor11 nfs: rpc.mountd startup succeeded
</snip>
This tells me that I probably should not do "nfs REstart" but "nfs start" in /etc/rc.d/rc.local but nothing more...
3. When I su and do this on the server itself :
-- mount 192.168.1.11:/misc tmp/
it takes long (almost a minute) but it works
If I try to mount the same share from a client now I get a
"mount: RPC: Timed out"
4. When I do then (on the server)
-- umount tmp/
It again takes quite a while and I get this:
"Bad UMNT PBC: RPC: Timed out"
but the mount is gone! (so far so good)
5. When I try to remount (on the server) the same way as in 3. I get after apropriate waiting time again a
"RPC: RPC Timed out"
If I restart the nfs-server (takes long but works) I can mount the share again on the client or on the server but only on one machine at a time.
Grateful for advice!
Tassilo