RE: Applying CPU to RAC with physical standby
Date: Wed, 28 May 2008 08:44:38 -0700
Metalink note 278641.1 gives instructions for applying a patchset with a 10g physical standby in place and, although I haven't tried it in a RAC environment, only with standby, but I suspect it could be extrapolated. Shut down the primary (in this case the RAC node which is the primary) and the standby, install the patch and then with both the primary and standby open run catpatch, then follow the instructions for a rolling upgrade on other nodes? I've not tried it but it seems like it should work.
Univ. of California at Davis
IET Campus Data Center
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of S. Anthony Sequeira Sent: Wednesday, May 28, 2008 5:13 AM
To: Oracle List
Subject: Re: Applying CPU to RAC with physical standby
On Wed, 2008-05-28 at 12:37 +0100, jason arneil wrote:
> I think you should be OK to apply the catpatch.sql with the database
> up, the patch instructions are clear on that. I don't think having
> dataguard involved impacts on this, why should it?
Yes, but the standby patch application note (187242.1) specifically mentions that all nodes must be down while the entire patch is being applied. This is what I'm querying with Oracle Support, and the list. I too believe that having a standby should not make a difference, but apparently it does.
> You must however
> start the database in upgrade mode to perform the view recompile -
> you know this already!
> At this stage I can't see how you can have a node open for business
> except if you want to offer your users some read only activity, then
> you could use your standby read only while you were doing the view
We have scheduled downtime in a short window between batch runs on a Sunday afternoon. Oracle Support are dragging their heels on this, if I have to take the database down for an extended period of time, it has to go back to the business.
> This though is far removed from the marketing hype of "rolling
Agreed, this begs the question, why use RAC and pay through the nose? But that's probably been gone through many times, and is not really relevant to the thread.
-- S. Anthony Sequeira ++ At the source of every error which is blamed on the computer you will find at least two human errors, including the error of blaming it on the computer. ++ -- http://www.freelists.org/webpage/oracle-l -- http://www.freelists.org/webpage/oracle-lReceived on Wed May 28 2008 - 10:44:38 CDT