Veritas-bu

[Veritas-bu] Conditional Expiration

2000-10-12 19:21:15
Subject: [Veritas-bu] Conditional Expiration
From: Joshua Fielden jfielden AT excitecorp DOT com
Date: Thu, 12 Oct 2000 16:21:15 -0700
Actually, I am scheduling a meeting with NetB project management and 
engineering on this very same issue. It does NOT do conditional expires. Legato 
and EDM do. Our estimate is this will cost us ~$250k in tapes alone in order to 
keep retentions on all fulls one week longer than the 
incrementals/differentials. This is not counting admin time to catch these 
issues, or write complex scripts to make sure data loss does not occur. (yes, 
that much. We have 14k active tapes right now) I am thinking of taking a 
ccouple of weeks, clearing my schedule, and writing this feature on top of the 
product with scripting using medialist and expdate, etc.

As far as I am concerned, this is a design flaw, and is treated as such in all 
my contacts with Veritas. Call this in to support as an RFE. 

In my opinion, the entire scheduler for this product was written by people who 
haven't actually performed an enterprise backup in the last 5 years, if ever. 
This is a huge sticking point for us in all it's many-faceted broken glory.

JF

On Thu, Oct 12, 2000 at 06:08:21PM -0500, John_Wang AT enron DOT net filled up 
my inbox with:
> 
> Hello
> 
> Is it possible to structure conditional expirations?
> 
> ie.: Don't expire a full backup till at least another full backup (or two) has
> been generated.  Expire the cumulatives when the underlying full backup has 
> been
> expired, expire the differential when the prior differential, cumulative or 
> full
> has been expired.
> 
> Time based expirations tend to not handle exceptions well ie.: if a full 
> backup
> gets preempted for some reason, the prior full backup will still time out.
> Also time based expirations tend to result in leftovers that don't need to be
> kept around ie.: when a full backup expires, the subsequent cumulatives and
> differentials between that full backup and a more recent one is insufficient 
> to
> guarantee recovery of a filesystem state so you might as well not have them or
> you may be setting unrealistic expectations.
> 
> I apologize if this is beyond the scope of the product but policy based
> expiration was the basis of my old home grown backup scripts back in the early
> 90's several jobs ago and I don't like surprises just because a backup got
> preempted.
> 
> Regards,
> John I Wang
> Sr. Systems Engineer
> Steverson Information Professionals
> 
> ---
> Enron Broadband Services
> 3 Allen Center, Room 337C
> PH (713) 345-6238
> 
> 
> _______________________________________________
> Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu

-- 
"Any man page that includes the words "USE AT YOUR OWN RISK. BEWARE OF DOG. 
SLIPPERY WHEN WET" means trouble" - Michael Lucas
Joshua Fielden, Senior Systems Administrator and Backups Team Lead
eXcite@Home, Inc. jfielden AT excitecorp DOT com 650-556-3316



<Prev in Thread] Current Thread [Next in Thread>