Amanda-Users

Re: NFS mount as second holding disk

2003-10-24 10:57:21
Subject: Re: NFS mount as second holding disk
From: Gene Heskett <gene.heskett AT verizon DOT net>
To: Hans Kinwel <hansk AT cistron DOT nl>, amanda-users AT amanda DOT org
Date: Fri, 24 Oct 2003 10:50:17 -0400
On Friday 24 October 2003 10:06, Hans Kinwel wrote:
>On Thu, Oct 23, 2003 at 11:05:57PM -0400, Gene Heskett wrote:
>> >> Generally speaking, dump is not the prefered utility for use
>> >> with amanda.  We seem to have gradually come to prefer tar, in
>> >> any version 1.13-19 or higher.
>
>and
>
>> Most of the comments I've read on this list from folks who have
>> run dump, seem to be less than enthusiastic after it bites them in
>> a recovery attempt.
>
>and more.
>
>Personally, I am a content dump user and I never had any problem
>recovering.  As far as I remember I have not seen any, or more than
> tar, failed recovery attempts on this mailinglist.
>
>Last week I decided to find out about the dump scare, as I seemed to
>remember that someone on this list said that dump is more or less as
>safe as tar is, given normal usage.
>
>Despite all my searching, I couldn't find that message on the list
> again. However, I found something better: the official word from
> the dump authors.
>
>Here it is.  http://dump.sourceforge.net/isdumpdeprecated.html
>
>I encourage anyone interested to read it.

Thanks for that bit of reasearch, Hans.

And it does seem to be an unvarnished attempt to clarify the problems 
and pitfalls of both approaches.  If one can re-mount the filesystem 
ro for the duration of the dump, this would seem to alleviate the 
major headache.  Or one can make a snapshot, but would not that 
snapshot take up as much space as the original?  And how long would 
it take to do a snapshot on a 40gig filesystem?

In re tars mods to atime.  I haven't gotten into a situation that 
requires I survey that except when we were cleaning up a rootkit some 
years ago.  Objections to that might be fixed by having a different 
open vector that was only available to tar and its ilk, and which 
didn't modify the atime.  Because of the time involved, and the 
cacheing in the normal access, restoreing the atime doesn't seem to 
me to be a wholly satisfactory solution, so the true bypassing would 
seem to be the only solution for that.

As far as "file changed while we read it" messages from tar, I've seen 
a bunch of them over the years but I have not yet been bit by 
recovering a bad file that was important enough to make me mutter 
about the authors geneology.  I suppose that day will happen, Murphy 
guarantees it will eventually.

I agree, everyone should read it, then make thier own mind up and live 
with it.

At this point, and because I can break disklist entries up into 
manageable pieces, I'll continue to use tar.  To use dump, I'd have 
to start by buying a tape drive about 10x bigger.  Nah, not for a 
home user, its not economically justifiable when I'm retired, meaning 
I really should be watching how I spend my money.


-- 
Cheers, Gene
AMD K6-III@500mhz 320M
Athlon1600XP@1400mhz  512M
99.27% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com attornies please note, additions to this message
by Gene Heskett are:
Copyright 2003 by Maurice Eugene Heskett, all rights reserved.