Re: rac vs dataguard
Date: Thu, 16 Jul 2009 20:41:54 +0200
Some other things to consider that have not yet been mentioned
- You can turn on flashback database for your physical standby, and activate it to do your reporting or do tests (that write to the database), or you can open it read only. For example you can do this during office hours or at predefined times. Then, later on you flash it back and put it back into recovery mode during the night. In the event of failure of the primary site, you will have to apply all logs to bring the standby up to date before you can switch over. So there's a tradeoff of using the resources for reporting vs HA (switchover takes longer)
- With 11g, you can even open the physical standby read only while it is in recovery mode (take a look at the active data guard option) -- for some environments, this may be reason enough to upgrade, as contrary to the physical standby that you only open read only at a given point in time (and therefore start lagging behind the primary) this is a near-realtime reporting solution.
Just my CHF0.02 :-)
Stefan P Knecht
CEO & Founder
10046 Consulting GmbH
Cell +41 (0) 79 571 36 27
On Thu, Jul 16, 2009 at 1:44 PM, ed lewis <eglewis71_at_gmail.com> wrote:
> We are planning to implement a 2 - node rac cluster, EE 10.2.0.4,
> solaris 10,
> using ASM.
> The plan is to also use dataguard to create a logical standby db for
> I was considering the idea of creating a 3rd member of the rac cluster,
> dedicate it to reporting, instead of using dataguard. There would be the
> cost for rac, but the adminstration may be simpler.
> Has anyone done this ? What are the pros and cons ?
> Thanks for your feedback.
> ed lewis