Re: Streams Instantiation Using RMAN Convert Database - 10g

thanks for the reply. right, so some of the uptime that is read only is during 
the rman convert process, which will have to do the following in my case:
1. pull files out of asm
2. convert undo and system and pull out of asm

I wonder if the convert process - which is basically transportable database, 
can instead be run on a physical standby while the source itself remains open 
Instead of rman convert, I could initially run the rman duplicate (which 
doesn't require the source to be read only), once the duplicate completes, run 
the rman convert on the duplicate database and use that to instantiate the 
final destination/target database. Sure it needs to be tested, but does it 
sound feasible?

From: Mark W. Farnham <mwf@xxxxxxxx>
To: oracledbawannabe@xxxxxxxxx; oracle-l@xxxxxxxxxxxxx
Sent: Sun, May 16, 2010 3:57:32 PM
Subject: RE: Streams Instantiation Using RMAN Convert Database - 10g

There is little to no downtime, but some
of the up time is read only. Only use this method if that usage pattern fits
your needs.

From:oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On 
Behalf Of Oracle Dba Wannabe
Sent: Sunday, May 16, 2010 6:11 AM
To: oracledbawannabe@xxxxxxxxx;
Subject: Re: Streams Instantiation
Using RMAN Convert Database - 10g
Or perhaps the source database remains online/available to users during
the 'transfer of files' from one platform to another - but remains offline
during the rman convert command - even so, running the rman convert command on
the source would still lead to considerable downtime.
Sorry I forgot to mention this for a migration from solaris (sparc) to hp-ux
(itanium) - both big endian.

From:Oracle Dba
Wannabe <oracledbawannabe@xxxxxxxxx>
To: oracle-l@xxxxxxxxxxxxx
Sent: Sun, May 16, 2010 3:02:44 PM
Subject: Streams Instantiation
Using RMAN Convert Database - 10g

Hi, 3 Node RAC, ASM - Source Database.
Has anyone used RMAN Convert as an instantiation option in streams?, Im getting
conflicting information, or perhaps misunderstanding what is written in the
The Streams Replication Guide under the 'Performing Instantiations' chapter,
appears to imply that the source database can continue to be up during this
process. However appendix c, which talks about using this method states that
the source database needs to be open read only. If this is the case I fail to
see how this process achieves little to no downtime - since appendix c appears
to talk about a source conversion -even though I don't need to convert all
datafiles I still need to pull them out of ASM. For a large database that can
take some time.
If anyone out there has used this process I'd like to hear your thoughts or
experience with it?......Thanks


Other related posts: