Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site uwvax.ARPA
Path: utzoo!watmath!clyde!burl!ulysses!mhuxl!houxm!houxz!vax135!cornell!uw-beaver!tektronix!hplabs!hao!seismo!uwvax!dave
From: dave@uwvax.ARPA
Newsgroups: net.unix-wizards
Subject: Re: using tar for incremental backups on 4.2 systems
Message-ID: <342@uwvax.ARPA>
Date: Thu, 19-Jul-84 09:44:38 EDT
Article-I.D.: uwvax.342
Posted: Thu Jul 19 09:44:38 1984
Date-Received: Sat, 21-Jul-84 03:11:04 EDT
References: <2085@sri-arpa.UUCP>
Organization: U of Wisconsin CS Dept
Lines: 19

We have used tar (or a subset of tar actually, it generates tar-tapes)
for incrementals since 4.1 times.  The only problem with the scheme
suggested is the 'find'.  Find is amazingly slow!  I wrote a new find
program which runs through the inode list on the filesystems to be
dumped (gleened from fstab), makes a bitmap of the files to be dumped
and then does a fast transversal of the directory structure for the
names ('sounds like dump' you say?  Yes, that's where I got the idea).
This takes about 10 mins when searching through 1++ gig.  The
differences to tar are that it reads the files it dumps from a file and
can dump to remote tapedrives (only under 4.2).  We also made a
complementary restore program which gives the ability to restore a file
to places other than where they originated.  This all runs very well
(except when the tapedrives die).

Now if I could only get the incremental dump history to work right.....
-- 
Dave Cohrs @ wisconsin
...!{allegra,heurikon,ihnp4,seismo,ucbvax,uwm-evax}!uwvax!dave
dave@wisc-rsch.arpa