> -----Original Message-----
> From: Sean Noonan
> Sent: 25 July 2006 22:45
> 
> <snip>
> > 
> > Use the -h option of amrestore to confirm if it was compressed,
> > probably with gzip.
> > 
> > If that is the case you will have to put a gzip -d command in
> > the pipeline between amrestore and restore.

My understanding is that amrestore automatically does any uncompression 
necessary so the explicit gzip -d isn't needed.

> Thanks for the reply.  I'm quite sure it was gzip compressed, 
> the amanda
> config files specifies "compress client fast" for this 
> volume.  Still no love, though.  Here's what I got:
> 
> freebee# amrestore -p /dev/nsa0 freebee aacd0s1f | gzip -d | restore -ivf -
> Verify tape and initialize maps
> amrestore: missing file header block
> amrestore:   2: skipping freebee.aacd0s1e.20060721.0
> amrestore:  10: reached end of information
> 
> gzip: stdin: unexpected end of file
> End-of-tape encountered
> Tape is not a dump tape

That looks like amrestore isn't finding the dump file on the tape and is 
running off the end of tape.

I'm puzzled by the "missing file header block".   That implies that the tape 
isn't positioned at the start of a tape file when you run amrestore.  What does 
"mt stat" say before you run amrestore?  Is the tape positioned at the very 
start of the tape file you want (or at least somewhere before there)?

Is that all the output you get from amrestore?  Does it really skip from 2: to 
10:?

Paul