Re: Active Dataguard -- Cascade Standby or not
Date: Wed, 7 Mar 2012 16:35:41 -0800 (PST)
You are welcome! You have broached a very interesting subject which was discussed very intently during our project. The long-term goal was to have the Observer installed on an independent data center (C) different from the PRIMARY (Data Center A) and HA (Data Center B). But Data Center C is not yet available. In the interim, we will have the Observer running on A with the option to run on B (if and when necessary) similar to an Failover Cluster configuration. We have Hitachi True Copy replicating the all important fsfo*.dat file (key to Observer startup) from A to B. This replication is done in synchronous mode, such that if there is an Observer failure on A, B has the accurate version/copy of the fsfo*.dat file for the Observer startup. Again, this was done keeping in mind the customer's all important requirement - Automation.
This topic of Observer configuration and setup in itself can be a discussion that can last many days :))) And there are different failure scenarios that one needs to be understand, plan for and deal with. If you have not already done so, I'd happily recommend that you read a presentation by Martin Bracher from Trivadis (albeit from 2008 but still very relevant). Martin has done an excellent job on the subject, kudos to him!
Phone - +1-650-743-6060
http://www.linkedin.com/in/gajakrishnavaidyanathaCo-author:Oracle Insights:Tales of the Oak Table - http://www.apress.com/book/bookDisplay.html?bID14 Co-author:Oracle Performance Tuning 101 - http://www.amazon.com/gp/reader/0072131454/ref=sib_dp_pt/102-6130796-4625766 Enabling Cloud Deployment & Management for Oracle Databases
From: Stalin <stalinsk_at_gmail.com>
To: Gaja Krishna Vaidyanatha <gajav_at_yahoo.com> Cc: Oracle-L List <oracle-l_at_freelists.org> Sent: Wednesday, March 7, 2012 3:53 PM
Subject: Re: Active Dataguard -- Cascade Standby or not
Thanks Gaja for sharing us the experiences from your recent POC. I appreciate it.
Apparently our project ended-up with having only One ADG and One DR side off of primary, similar to your setup except primary being RAC.
Just curious on observer setup. Was observer installed on a separate site different from primary and HA or monitored from primary always? How did you establish HA for observer process as there can be only be one observer monitoring the health of DG setup. Is this managed by GC.