Path: utzoo!attcan!uunet!samsung!zaphod.mps.ohio-state.edu!math.lsa.umich.edu!emv
From: emv@math.lsa.umich.edu (Edward Vielmetti)
Newsgroups: comp.unix.ultrix
Subject: Re: dumping active file systems
Message-ID: <EMV.90Mar20145054@duby.math.lsa.umich.edu>
Date: 20 Mar 90 19:50:54 GMT
References: <384@vela.acs.oakland.edu>
Sender: news@math.lsa.umich.edu
Organization: University of Michigan Math Dept., Ann Arbor MI.
Lines: 17
In-reply-to: schemers@vela.acs.oakland.edu's message of 17 Mar 90 02:37:23 GMT


My understanding is that the Ultrix ``dump'' has had some of the
Purdue modifications into it to harden it against problems with
dumping live filesystems.  This information courtesy of Rich Kulawiec,
partial author of some of the Purdue changes, who says he's using the
supplied dump on the DS3100.

Here's a worst case test for dump, courtesy of 
Don Speck <mangler@csvax.caltech.edu>:

    mkdir foo; cd foo; sleep 999999 & cd ..; rmdir foo
    dump that filesystem
    restore that filesystem someplace (e.g. /usr/tmp)

see if restore dumps core on this one.

--Ed