Questions about HUP'ing nfsd

[James]

On Fri, 2007-10-12 at 08:25 -0700, Michael Smith wrote:

> Hello All:
> 
> We have a NAS that's running 6.2 with nfds, mountd, rpc_statd,  
> rpcbind and rpc_lockd.  Last night we had a scenario where nfs  
> clients, once disconnected, couldn't reconnect to the NAS, reporting  
> RPC timeouts.
> 

I've had RPC timeouts before. Turned out my NFS was misconfigured - I
had a weird flag in /etc/rc.conf that was preventing mountd from
loading. Could be someone made an undocumented change - you may want to
check it.




> My question is, in troubleshooting this sort of thing, is there a  
> proper sequence for stopping and restarting the various services  
> associated with nfs?   Any hints would be greatly appreciated.
> 



Linux usually likes an exportfs -r, service portmap restart, service nfs restart. I've usually gone for a 
similar thing in FreeBSD. Something like:


/etc/rc.d/mountd onereload
/etc/rc.d/nfs restart



I think that does it all. Otherwise, there's always shutdown NOW and
then a ctrl-d. If you're not sshing in, of course.

James