Comments on: RMAN Redundancy is not a Viable Retention Policy

By: Ian Murphy

Ian Murphy — Thu, 22 Dec 2011 13:51:45 +0000

Late to the party but in answer to Michael B above, RMAN should *not* delete any backup files (backups of Archive Redo Logs, data files, control files etc) that would be required to meet your “recovery window”.

For example, if you have a recovery window of 3 days and your last good backup of the datafiles was a week ago, this backup will not be deleted (ever) until you’ve got another one that is older than 3 days but newer than the previous one.

Make sense?

Cheers,

Ian

By: Michael B.

Michael B. — Wed, 09 Nov 2011 21:37:31 +0000

I am starting to use recovery window instead of redundancy for my RMAN backups but I have a concern:

Every once in a while my backups will fail and it takes a couple of days for a technician to notice this. If my recovery window is 2 days and I don’t take a backup for 3 days then it will automatically delete my only good backup!!! Is this incorrect? Will it have the intelligence not to delete the last backup?

By: rman

rman — Sat, 11 Sep 2010 22:46:17 +0000

Thank you so much for this precious information.

By: zhix

zhix — Sat, 24 Apr 2010 22:05:57 +0000

what about – backup validate check logical …?

By: RMAN Redundancy is not a Viable Retention Policy « die Seilerwerks

RMAN Redundancy is not a Viable Retention Policy « die Seilerwerks — Wed, 19 Aug 2009 18:17:52 +0000

[...] leave a comment » Originally posted on The Pythian Group blog. [...]

By: Log Buffer #138: A Carnival of the Vanities for DBAs

Log Buffer #138: A Carnival of the Vanities for DBAs — Fri, 06 Mar 2009 17:05:17 +0000

[...] Having covered IBM, let’s switch to Oracle and Eric Emrick, where he talks about database continuity for a while, and then let’s go over to Pythian’s Don Seiler and his talk on how RMAN isn’t good for retention policies, in his blog RMAN Redundancy is not a Viable Policy. [...]

By: Chris_c

Chris_c — Thu, 05 Mar 2009 14:17:21 +0000

I’ve never liked redundancy based retention, recovery window always seemed more precise, although there are a couple of oddities here My understanding is that the datafile backup should have failed when it found the corruption unless configured to ignore it, also having the backupsets contain one file each seems to be required.

I’ve seen the tape backup thing as well working at a site with a 4 day SLA to fix tape failures and a three day archivelog disk retention policy, has some interesting conversations with the storage guys on that one.

By: Bradd Piontek

Bradd Piontek — Wed, 04 Mar 2009 22:57:12 +0000

Memories , like the corner of my mind.
Misty water-colored memories.
Of the way we were
Scattered pictures,
Of the smiles we left behind
Smiles we gave to one another
For the way we were
Can it be that it was all so simple then?
Or has time re-written every line?
If we had the chance to do it all again
Tell me, would we? could we?

No. we definitely could not, should not do it all again.