Subject: Re: Errors reading dump on Exabyte 8200
To: Claude Marinier <claude.marinier@dreo.dnd.ca>
From: Robert Alexander Baxter <alex@santa.asf.alaska.edu>
List: netbsd-help
Date: 08/28/1998 08:33:05
Marinier, M. Claude, G. wrote:
> 
> I am setting up a server: NetBSD 1.3.2 (GENERIC for now) on a Pentium II
> 233 MHz with 128 MB, an AHA-2940, and an Exabyte 8200 (also tried a Python
> DAT). I have created a dump tape on the 8200 but cannot read it (restore
> -i does not get to the prompt). There were some I/O errors while the dump
> script ran. I switched to a DAT drive and got the same kind of behaviour.
> However, I was able to write to it and read from it without any errors
> using 'tar cf' and 'tar tf'.
> 
> I have searched for an answer but found none. I get the impression that
> this have been done many times before. I must be missing something basic.
> I have included the dump script.
> 

I can't help you with the Python DAT, but as for the Exabyte 8200,
I have one too, and in my experience it is best not to use any blocksize
other than 1024 *bytes* (2 512 byte blocks for UNIX commands).  Some
kinds
of tar will determine the block size to use to read the tape.  Some
kinds of tar will not be able to read tar tapes you make, unless you
specified the block size when you wrote the tape, and then give tar that
block size information when you read the tape.

I don't know how the dump command handles possibly different
block sizes.

I'm not positive, but it looks like in your dump script you are using
a block size of 126 512 byte blocks (64,512 bytes per block).  Try
changing this number to 2 instead of 126, and the Exabyte 8200 might
work better for you.

Hope this helps,
 -Alex (:-)
  alex@santa.asf.alaska.edu