amrecover: Can't read file header from remote host

Howdy,

Having a problem getting amrecover to work when restoring from a remote
host.  My backup server is running Redhat 7.3 with the included version of
amanda:

[cfuhrman@comanche tmp]$ rpm -qa | grep amanda
amanda-devel-2.4.2p2-7
amanda-server-2.4.2p2-7
amanda-2.4.2p2-7
amanda-client-2.4.2p2-7

The remote servers are running OpenBSD 3.2 with amanda compiled from the
ports tree:

bash-2.05b# pkg_info | grep amanda
amanda-client-2.4.2.2 network-capable tape backup (client only)

Here's what a typical session looks like when attempting to restore from a
remote client.  In this example, comanche is the amanda server and
algonquin is the amanda client.  I'm running from algonquin here:

bash-2.05b# /usr/local/sbin/amrecover -C WEEKLY -s comanche -t comanche -d
/dev/
ntape
AMRECOVER Version 2.4.2p2. Contacting server on comanche ...
220 comanche AMANDA index server (2.4.2p2) ready.
200 Access OK
Setting restore date to today (2003-03-15)
200 Working date set to 2003-03-15.
200 Config set to WEEKLY.
501 No index records for host: algonquin.beartribe.org. Invalid?
Trying algonquin.beartribe.org ...
501 No index records for host: algonquin.beartribe.org. Invalid?
Trying algonquin ...
200 Dump host set to algonquin.
$CWD '/home/cfuhrman' is on disk '/home' mounted at '/home'.
200 Disk set to /home.
/home/cfuhrman
WARNING: not on root of selected filesystem, check man-page!
amrecover> ls
amrecover> add hosts.old
Added /cfuhrman/hosts.old
amrecover> extract

Extracting files using tape drive /dev/ntape on host comanche.
The following tapes are needed: WEEKLY02

Restoring files into directory /home/cfuhrman
Continue? [Y/n]:

Load tape WEEKLY02 now
Continue? [Y/n]:
EOF, check amidxtaped.debug file on comanche.
amrecover: short block 0 bytes
UNKNOWN file
amrecover: Can't read file header
extract_list - child returned non-zero status: 1
Continue? [Y/n]: n
amrecover> exit
200 Good bye.

And here is what my amidxtaped.debug file contained.  No indications of a
problem:

amidxtaped: debug 1 pid 5712 ruid 33 euid 33 start time Sat Mar 15 05:40:36 2003
amidxtaped: version 2.4.2p2
> SECURITY USER root
bsd security: remote host algonquin.beartribe.org user root local user amanda
amandahosts security check passed
> 6
amrestore_nargs=6
> -h
> -p
> /dev/ntape
> algonquin
> ^/home$
> 20030314
Ready to execv amrestore with:
path = /usr/sbin/amrestore
argv[0] = "amrestore"
argv[1] = "-h"
argv[2] = "-p"
argv[3] = "/dev/ntape"
argv[4] = "algonquin"
argv[5] = "^/home$"
argv[6] = "20030314"
amrestore:   0: skipping start of tape: date 20030314 label WEEKLY02
amrestore:   1: skipping comanche._export0_pub.20030314.1
amidxtaped: amrestore terminated normally with status: 0
Rewinding tape: done
amidxtaped: pid 5712 finish time Sat Mar 15 05:41:01 2003

Now what's weird is that if I try to restore from the server machine, then
amrecover works perfectly okay.  It just gives the above behavior when
attempting to restore from a client.  Using mt -f /dev/ntape fsf 1 and dd
to extract from tape manually works perfectly okay.  For all filesystems
I'm using GnuTAR (/usr/bin/tar on the Linux box, /usr/local/bin/gtar on
the OpenBSD systems).

And yes, I rewound the tape before beginning :)

I'm stuck here so if anyone can point in me the right direction, I'd
greatly appreciate it!

Cheers!

-- 
Chris Fuhrman
cfuhrman AT pobox DOT com

http://www.pobox.com/~cfuhrman