re: HACMP and Enterprise Server

From: Sanchia Baker <sabaker_at_seeboard.com>
Date: 1998/04/20
Message-ID: <199804200939.AA01938_at_interlock.csw.com>#1/1


We use Oracle V 7.3.4 in a two node HACMP , standby envirnoment It works okay here, but what we did was set up the Oracle application ser ver closedown script to rcp a flag file over to the standby node on fail ure, if the closdown of the oracle databases was successful on the primar y node. The standby node then checks for the presence of this file, after  the takeover of the resource group and before starting up the oracle dat abases.
If this file exists, it assumes that oracle was closedown properly and starts the databases and removes the flag file. If it doesn't exist, it sends a message to the ITO message browser advising that a DBA be contact ed urgently and the databases are not brought on -line.

A DBA would then check out the database and bring it on-line when they've  okayed it.

Sanchia
Disclaimer: My opinions are my own. Received on Mon Apr 20 1998 - 00:00:00 CEST

Original text of this message