I got one response in this, from "Paul M. Aoki" <aoki_at_CS.Berkeley.EDU>,
who has had the same problem with 3.2.
I got my final solution my calling Dec and getting a mega patch set for
OSF/1 3.2 which included a patch for this problem.
Here is Paul M. Aoki message and my reply, which describes the patch I
applied :
---------------------------------------------------------------
>From valerie_at_cs.umass.edu Wed Jun 17 10:23:26 1998
Date: Wed, 17 Jun 1998 09:39:35 -0400 (EDT)
From: Valerie Caro <valerie_at_cs.umass.edu>
To: "Paul M. Aoki" <aoki_at_CS.Berkeley.EDU>
Cc: Valerie Caro <valerie_at_cs.umass.edu>
Subject: Re: Problem with Alpha crashing with dupdone: bad copy
On Wed, 17 Jun 1998, Paul M. Aoki wrote:
Thanks for the info. I had also found that there was a
patch for later versions of the OS, but I can't update at this time
unfortunately. I ended up calling Dec support and they put a mega
patch for 3.2 in an ftp area for me. I found the patch I needed in there.
I think it is this one but I applied about 20 different patches from
the set, so I am not sure :
-------------
/usr/sys/BINARY/nfs_server.o
CHECKSUM: 20710 300 RCS: 1.1.76.4 (nfs_server.c)
/usr/sys/BINARY/nfs3_server.o
CHECKSUM: 40669 301 RCS: 1.1.29.2 (nfs3_server.c)
/usr/sys/BINARY/nfs_xdr.o
CHECKSUM: 55274 32 RCS: 1.1.30.2 (nfs_xdr.c)
---------------------
PATCH ID: OSF320-164
Supersedes Patch: OSF320-104, OSF320-126
PROBLEM 3: (Patch ID: OSF320-164) (HPAQ72013)
**********
System hang or system panic with "kernel memory fault" if an nfs server
is given corrupted data.
---------------------------------------------
> fyi, if you are running DU 3.2C, you can get a megapatch from
> www.service.digital.com:8031 that contains a successor to this patch.
> (dig around for 350-020 using their patch search engine.)
>
> we have some systems at DU 3.2 as well. same thing, you get almost to
> the end and whammo, the machine goes down with that "svcudp_dupdone"
> message. the only heuristic "fix" i've found for this is to shut off
> our solaris clients while the machine reboots. how's that for something
> you really want to do after a midnight power failure.
>
> as an added incentive to upgrade, i've also found that DU 3.2 can be
> crashed by x86 solaris 2.5.1 clients doing a lot of NFS3 file operations
> (e.g., those done by netscape while managing its cache directories).
> even better, our x86 solaris 2.5.1 clients can crash DU 3.2 just by
> sending NFS2 mount requests.
> --
> Paul M. Aoki | University of California at Berkeley
> aoki_at_CS.Berkeley.EDU | Dept. of EECS, Computer Science Division #1776
> | Berkeley, CA 94720-1776
>
Original problem:
----------------------------------------------------------------------
Date: Tue, 16 Jun 1998 11:48:40 -0400 (EDT)
From: Valerie Caro <valerie_at_cs.umass.edu>
To: alpha-osf-managers_at_ornl.gov
Cc: Valerie Caro <valerie_at_cs.umass.edu>
Subject: Problem with Alpha crashing with dupdone: bad copy
I have Dec 3000/500 alpha running Digital Unix 3.2, which started having
a problem booting today after a power failure.
It seems to boot fine until it gets to the
nfs server programs (mountd ...). then it dumps with the error:
dupdone: bad copy 0c8d28ec00 0
panic(cpu 0): svcudp_dupdone
The only mention I can find of this errors indicates it could be:
The Patch ID is:
=======
PROBLEM: (Patch ID: OSF350-020) (HPAQ72013)
System hang or system panic with "kernel memory fault" if an nfs server
is given corrupted data.
=======
Since this patch is not For OSF3.2, and I cannot find the patch anyway,
it does not seem to apply to this situation. Does nayone know what this
error might be from, and what I need to do to fix it?
---
---
Valerie Caro Computer Science Computing Facility,
valerie_at_cs.umass.edu LGRC Room A313,
Tel: (413)545-4442 University of Massachusetts
Fax: (413)545-1249 Amherst, MA 01003
Received on Wed Jun 17 1998 - 16:32:10 NZST