Re: ****** RMAN backup for multi TB databases ******

  • From: Tim Gorman <tim@xxxxxxxxxxxxx>
  • To: <oracle-l@xxxxxxxxxxxxx>
  • Date: Sun, 22 Aug 2004 10:17:46 -0600

Obviously, it depends on what needs to be restored - one datafile, a couple
of them, or all of them.  The level-0 backup of 13Tb is approaching 24
hours, so that should provide perspective.

However, just by sheer odds based on space consumption, it is the READ ONLY
tablespaces (which only need to be restored, not recovered) that are most
likely to get "hit" by some failure.  The odds of having to restore and
recover the entire 35Tb at once is pretty slim, and at that point, you have
to compare the cost of possibly having to restore/recover 35Tb against the
benefits of having the 35Tb at all.  What would be the advantages and
disadvantages of throwing some of it away on purpose?  And, would it be
necessary to have a complete service outage for the entire restore/recover
operation, or can some of the database (i.e. tablespaces with "newer" data)
be up and running while the rest of the database is restored and recovered?

Nice thing about RMAN is that mistakes on restore are fewer and more

on 8/20/04 9:56 AM, DENNIS WILLIAMS at DWILLIAMS@xxxxxxxxxxxxx wrote:

> Tim - How long do they estimate the recovery time would be?
> Dennis Williams
> Lifetouch, Inc.
> "We all want progress, but if you're on the wrong road, progress means
> doing an about-turn and walking back to the right road; in that case,
> the man who turns back soonest is the most progressive."
> -- C.S. Lewis
> -----Original Message-----
> From: oracle-l-bounce@xxxxxxxxxxxxx
> [mailto:oracle-l-bounce@xxxxxxxxxxxxx]On Behalf Of Tim Gorman
> Sent: Thursday, August 19, 2004 11:06 PM
> To: oracle-l@xxxxxxxxxxxxx
> Subject: Re: ****** RMAN backup for multi TB databases ******
> I recently visited a shop where they use RMAN on a 35Tb DW.  22Tb are READ
> ONLY tablespaces, so their weekly level-0 is about 13Tb.  Their daily
> level-1 and level-2 average about 2-4Tb each, couple hundred Gb of
> archivelogs per day.  They run big-iron Solaris and Veritas Netbackup to Sun
> DLT tape drives (16 drives, I think).
> They're happy with RMAN, although that weekly level-0 is exceeding the 24hr
> backup window they've imposed.  I suggested that the level-0's be
> distributed across each day of the 7-day week, to even out their spikes;
> don't know if they implemented it or not.
> Also suggested that they look into table compression, but that will be a
> longer term thing.  At the moment, I think they opted to order 6-8 more tape
> drives...
> RMAN isn't a problem;  it's a big part of the solution...
> on 8/19/04 4:56 PM, Sam Reddy at iamanoracledba@xxxxxxxxx wrote:
>> Hi All,
>> Could you please share your experience if you have implemented RMAN backup
> on
>> a 5TB (Tera Byte) system ? I am concerned about backup time with
>> full/cumulative/ incremental backups when RMAN is used in conjunction with
>> Legato networker. Would you please send me if you have any white papers ?
>> Hardware setup at your shop and any other technical considerations ?
>> I appreciate your help.
>> Have a wonderful day!
>> Regards,
>> Sam.

Please see the official ORACLE-L FAQ:
To unsubscribe send email to:  oracle-l-request@xxxxxxxxxxxxx
put 'unsubscribe' in the subject line.
Archives are at
FAQ is at

Other related posts: