>ELF 32-bit MSB core file SPARC Version 1, from 'ufsrestore
So this isn't an Amanda problem, but a Solaris one.
Here are some thoughts:
* Go to Sun and look for any patches to ufsdump/ufsrestore, i.e.
make sure you're running the latest version.
* Reload the image to disk by hand and run ufsrestore against it
directly instead of in a pipeline:
cd /some/place
mt -f /dev/rmt/0cbn
amrestore /dev/rmt/0cbn admin1.corp.walid.com sda10
cd /where/ever
ufsrestore if /some/place/admin1.corp.walid.com.sda10.20010519.0
* If that doesn't work, see if any of your other systems are running
a different version of ufsrestore (either older or newer) and either
try the image there or copy the ufsrestore program over to your tape
server (use some temporary name) and try it there.
* If that doesn't work, try dd'ing the image to a **scratch** tape and
handing that to ufsrestore:
cd /some/place
mt -f /dev/rmt/0cbn
dd if=admin1.corp.walid.com.sda10.20010519.0 of=/dev/rmt/0cbn bs=32k
mt -f /dev/rmt/0cbn
cd /where/ever
ufsrestore ibf 64 /dev/rmt/0cbn
* If that still doesn't work, ask me offline of the list and I'll dig
into my dirty little bag of tricks for some other slight of hand.
One question first, what version of Solaris are you using?
BTW, at some point you'll want to change to using /dev/rmt/0cn instead
of /dev/rmt/0cbn. The 'b' gives you "BSD semantics", which can sometimes
cause problems with Amanda. Nothing related to what your seeing at the
moment, but there are other issues.
John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]
