Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uunet!brunix!sgf
From: sgf@cs.brown.edu (Sam Fulcomer)
Newsgroups: comp.sys.sgi
Subject: Re: rtar, rdump on SGI's
Message-ID: <28319@brunix.UUCP>
Date: 7 Feb 90 16:18:37 GMT
References: <7032@sdcc6.ucsd.edu>
Sender: news@brunix.UUCP
Reply-To: sgf@cs.brown.edu (Sam Fulcomer)
Distribution: usa
Organization: Brown University Department of Computer Science
Lines: 21

In article <7032@sdcc6.ucsd.edu> pa1081@sdcc13.ucsd.edu (pa1081) writes:
>I would like to do central administration on our SGI's by backing
>them up and doing all tape operations through our alliant FX/80
>which has a 1/2" tape drive.  When I called sgi customer support,
>they informed me that this would not be possible (due to byte swaps
>and such).

Huh? Why? What do they think byteorder(3N) does? Accessing a remote bsd-tape
from an SGI is a lot more straightforward than reading random foreign tapes
on local drives.

eg,

tar cf - /unix | rsh sun dd of=/dev/rst0
rsh sun tf /dev/rst0
rsh sun dd if=/dev/rst0 | tar tf -
tar cf - /unix | rsh sun tar tf -

all work thanks to byteorder(3N).

What's the problem?