This is my second try posting this... I hope I am not screwing
something up...

Hi, thanks for the great program.  I am having a problem with
amrestore/amrecover.  When amdump goes off, everything is fine, and
the report says everything is fine.  When I try amrestore though, I
get:

[root@riesling holding_disk]# mt -f /dev/nst0 rewind [root@riesling
holding_disk]# amrestore /dev/nst0 
amrestore: 0: skipping start of tape: date 20030216 label yorkgroup05 
amrestore: error reading file header: Input/output error

So maybe the tape and/or drive is dirty/bad.  Not so.  If I do this:

$ mt -f /dev/nst0 rewind; mt -f /dev/nst0 fsf filemarker
$ dd if=/dev/nst0 skip1 bs=32k of=backup.tar.gz

I can read the tapes just fine.

Now this doesn't happen with every tape.  I just had one today that was fine.

What may be of interest to the developers is that dd prints out a
little error message:

dd: warning: working around lseek kernel bug for file (/dev/nst0)
  of mt_type=0x72 -- see <sys/mtio.h> for the list of types

So my theory is that dd knows about a kernel lseek bug that amanda
does not.  Maybe if some one in the know peeked at the dd source and
the amanda source they could sort this out.

I am using Red Hat 8.0, on a dual AMD workstation, a VXA-1 external
drive attached to a aic7892: Ultra160 Wide Channel A controller, Linux
kernel 2.4.19, amanda 2.4.2p2-9 (as packaged by Red Hat), and
fileutils 4.1.9-11 (as packaged by Red Hat).

Thanks.

-- 
--------------------------------
Kevin Range
range006 AT tc.umn DOT edu

281 Kolthoff: 612-625-6317
--------------------------------