[lug] Sun Client Redhat AS3 NFS Probs

[Jeffrey A. St. Pierre]

On Thu, 13 May 2004, Hugh Brown wrote:

> 
> D. Stimits post was that the error messages implicate rquotad and to
> check your quota definitions.  If you haven't set up quotas yet, then
> I'd turn off rquotad and solve the nfs problem (if it remains).  Once
> that is settled, then I would tackle getting rquotad configured and
> behaving correctly.
> 
> So, I'd recommend the following:
> 
> Turn off rquotad on the RH AS3 box.  
> Comment out the automap for the linux share on the solaris box and
> reload autofs/automount.
> Mount the linux share by hand on the solaris box and then check for
> errors.
> 
> If errors persist, see if it is only the specific solaris box or if it
> happens on another one.
> 
> If they don't persist, then re-enable the automount and test again.
> 
> If all good, then re-enable rquotad after configuring quotas and see
> what you get.
> 
> FWIW,
> 
> Hugh
> 


Thanks... I'm getting closer.  I can generate this error by
running 'quota -v' on Linux or Sun clients, so it is not a
Sun/Linux issue.  And I didn't have the filesystem automounted on
the linux box I was testing from, I mounted it by hand.  So that
eliminates a little of the mystery.  

I was thinking this error was generated from the client side
but appears in the server logs.

i.e. The error on the server:

May 14 10:03:44 <server> rpc.rquotad: Can't find filesystem
mountpoint for directory /export/rd02/diag
May 14 10:03:44 <server> rpc.rquotad: No correct mountpoint
specified.

This didn't make sense, because /export/rd02/diag does exist on
the server.   However, it doesn't on the client.  The client only
sees it as /mnt/diag.  So I tested this theory by creating a
mountpoint that looked the same as the one on the server and
changed my fstab.

I remounted <server>:/export/rd02/diag as
<client>:/export/rd02/diag and ran '/usr/bin/quota -v' again and
still got the error message on the server.

I'm begining to think this is a bug in AS3 that I should report
to redhat.  Anyone willing to second me on that yet?

Oh, BTW... I did get all the latest updates to Enterprise 3
that Redhat released yesterday, to include a new kernel and a new
nfs-utils package.  The errors still occur.

Any other ideas as to why this would occur?

-Jeff