Veritas-bu

Re: [Veritas-bu] Quarterly Backups and Calendar Schedule

2008-03-31 23:45:51
Subject: Re: [Veritas-bu] Quarterly Backups and Calendar Schedule
From: "Ed Wilts" <ewilts AT ewilts DOT org>
To: "A Darren Dunham" <ddunham AT taos DOT com>
Date: Mon, 31 Mar 2008 22:27:53 -0500
On Mon, Mar 31, 2008 at 7:16 PM, A Darren Dunham <ddunham AT taos DOT com> wrote:
On Mon, Mar 31, 2008 at 05:48:58PM -0400, Curtis Preston wrote:
> A 7 year old backup is a 7 year old backup.  It's not an archive, and
> 7 years from now when you're trying to use that 7 year old backup as
> an archive and trying to retrieve something from it (you retrieve from
> archives.  You restore from backups), you'll know what I meant.
> You'll rue the day you decided to keep a backup for 7 years.  That's
> all I'm going to say about that.

Not worrying about the initial issue, what is the biggest problem you
see with using the backups as archives?  I'm often dealing with 7 year
(or supposedly longer) retentions at clients.  What should I be prepared
for?

Off the top of my head:

What tape drives  did you have 7 years ago?  Do you still have them?  Are you sure you can actually read 7 year old tapes?  When did you last test?  What software wrote the tape?  Can it still read your current tapes?  Does the application that required the data still exist?  If so, does it still read and understand the same format?  What host had the financial data 7 years ago?  If you hired somebody today, would that person know?  How are you going to find the data you wrote? 

Go find an intern, point him/her at your documentation and request a (probably vague) restore of 6-year old data.  See how long it takes...  Yeah, I'd like to restore the financials.  I think the host might have been named cherry.  No wait, that was the dev system, try apple instead.  No wait, that was the stage system, try orange.  How bent out of shape are you going to be if you actually did a restore for what you thought was production but was actually staging data that might not have been accurate?  It might have been close - perhaps too close to tell the difference.  Will the binaries even run on the current hardware platforms?  Or are you going to restore 32-bit binaries onto a 64-bit platform?  How do you know the current 64-bit binaries behave the same way as the 32-bit ones did?  Are they bug-for-bug compatible?  Are you going to restore a system that gets immediately penetrated by virii that didn't exist back then?  Did any of the software require specialized dongles or keys?  I've seen parallel-port dongles recently and boy was it fun watching our Windows guys try to restore that...not... 

I know I have tapes in my closet and no tape drives to read them.  I know I have "archives" that are in a format the current application won't deal with properly.  With luck, I'll have an "experienced" person on staff that can remember how to convert that data so that the current system can re-process it.

We're still in the process of re-writing old SDLT tapes on to LTO-3 format tapes.  By the time we're done, we'll probably be on LTO-6.  Hey, I didn't say we were perfect :-)

Whenever I get a request for an infinite retention, I give the person a half-hour lecture on why they shouldn't ask for anything that obscene.  We're not manufacturing airplane parts folks (but perhaps you are!).

--
Ed Wilts, Mounds View, MN, USA
mailto:ewilts AT ewilts DOT org
_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu