NFS problems on a heterogeneous network (VAX <-> IPC)

Baldoni_DR at cc.curtin.edu.au Baldoni_DR at cc.curtin.edu.au
Sat Mar 16 08:44:00 AEST 1991


G'day all - this may have been discussed before but I'm having real
hassles, and I need a quick solution.

We've got a network comprising of two SPARC-1's (running 4.0.3), a SPARC-2
(running 4.1.1), two IPC's (running 4.1.1), a VAX-11/785 (running
Ultrix-32 V3.0 [Rev #64]) and several Silicon Graphics IRIS workstations
(running IRIX 3.3.2).

Just about everything is cross-mounting filesystems from somewhere!  The
links to/from the IRIS's are fine.  The link between the two IPC's is fine
as is the link between the two SPARC-1's.  The SPARC-2 is minding its own
business.

The problem is between the VAX and the IPC's.  Now, it will quite happily
mount/utilise the filesystems exported from the IPC's and they will quite
happily mount the VAX's exported filesystems.  But, after a short period
(the length of which appears to be indeterminate), I start getting
messages of the form:

        NFS [fsstat|getattr] failed for server cutmcvax:
        RPC: Authentication error

It further tells me:
        why = Invalid client credential

In desperation, I told the VAX to export the filesystems in question to
ANY host that asked for them.  And, I still saw this message!

If anybody out there knows what's going on, PLEASE let me know immediately
either at the addresses listed below or at:

        flint at cutmcvax.cs.curtin.edu.au         (preferred contact)

Thanks for any help.

Phone:    +61-9-351-7450                              | Daniel Baldoni
PSImail:  psi%050529452300070::TBALDONID              | Computing Science
Internet: TBALDONID at cc.curtin.edu.au                  | Curtin University
Bitnet:   TBALDONID at cc.curtin.edu.au@cunyvm.bitnet    | GPO Box U1987
UUCP:     uunet!munnari.oz!cc.curtin.edu.au!TBALDONID | Perth
                                                      | Western Australia, 6001



More information about the Comp.sys.sun mailing list