Path: utzoo!utgpu!cunews!bnrgate!brchh104!brchs1!bnr.ca!rice.edu!sun-spots-request
From: wolfgang@wsrcc.com (Wolfgang S. Rupprecht)
Newsgroups: comp.sys.sun
Subject: Re: Tar problems
Keywords: No Digest Subjects in Unmoderated Mode
Message-ID: <3605@brchh104.bnr.ca>
Date: 5 Jun 91 14:17:00 GMT
Sender: news@brchh104.bnr.ca
Organization: Sunspots, Pseudo-Unmoderated
Lines: 31
Approved: sun-spots@rice.edu
X-Original-Date:    Sat, 25 May 1991 20:41:07 GMT

paolucci@chaos.ame.nd.edu (Dr. Samuel Paolucci) writes:
>I have a QIC tape with important information in it.  This tape was
>write protected.  [...]  A couple of days ago I tried to get stuff
>from the tape and very quickly tar stops without retreiving anything.
>If I persist by using mt by skipping files, it tells me that I've
>reached the end of the tape.

Sun's tape SCSI driver is pretty eager to rewind and call the quits as
soon as it hits a bad block.  I tried to read a 5 year old QIC tape
with a bad oxide spot after a few megabytes.  The Archive Viper would
read along until it hit the bad block.  The tape drive would try to
re-read the block a few times and give up with an error.  At this
point the Sun device driver would basically over-react and set the
tape-block number to unknown and force a rewind on the next tape open.
There was no way that I could think of to skip over this bad section
and continue reading the next good block.

What I did is find another company's computer and try to read the tape
on that.  I eventually read my tape of on a BULL DPX-2 (68030 SysV
Unix box).  I just did several dd's and got two multi-megabyte files
and several zero-length files.  I spliced the two real files together
with 3 blocks worth of NULLs and fed the resulting file to tar.  I got
all the date back except for those 3 trashed blocks.  Too bad one
can't use a Sun to do this and has to use some other company's
computers.

-wolfgang
-- 
Wolfgang Rupprecht    wolfgang@wsrcc.com (or) uunet!wsrcc!wolfgang
Snail Mail Address:   Box 6524, Alexandria, VA 22306-0524