RMAN Myths

From: John Kanagaraj <john.kanagaraj_at_hds.com>
Date: Tue, 13 Dec 2005 15:21:31 -0800
Message-ID: <BEE6A332AA61424EAE305CF89D6F75C81E6EC7@USSCCEVS101.corp.hds.com>

Thought I would share this titbit, as I was researching for something else and stumbled onto this:

Apparently, configuration of the LARGE_POOL_SIZE in a RMAN backed up database is required *ONLY* if using DBWR_IO_SLAVES or BACKUP_TAPE_IO_SLAVES. If these parameters are not used, the Input and Output buffers for the RMAN process is used out of the session's PGA. (The Large Pool of course should be configured if MTS is in use, but it does not necessarily have to be configured for RMAN to be used).

This saves us from having to set/restart RMAN backed up databases for this parameter ;0) [For other RMAN myths, see ML Note 134214.1]

Incidentally, Myth # 2 has to be modified to add a caveat for 'block change tracking' enabled Oracle Database 10g databases - See ML Note 262853.1.

Regards,
John Kanagaraj

--
http://www.freelists.org/webpage/oracle-l

Received on Tue Dec 13 2005 - 17:21:42 CST