OPS Oddity

From: Kevin Brand <kbrand_at_gtetel.com>
Date: 1997/06/28
Message-ID: <01bc83cc$f6603b00$7c5c5b8f@netman.gtetel.com>#1/1

I'll be configuring a two node cluster of UNIX boxes with a shared disk sub-system running OPS ( as soon as I get the UNIX vendors DLM software ). The architecture of the application, which cannot change, does not allow a single Instance of Oracle to house more than one Company schema. So... in order to make available several differing data sets, I must provide multiple Instances within the cluster ( more than one per node ).

I understand that this is possible, but would like to hear of any experiences that someone else may have had. Any pitfalls in general? Any problems with running multiple copies of the DLM? Any PCM problems with multiple SGAs per node running OPS?

The platform: IBM RS/6000 R30x8, AIX4.1.4, HACMP 4.?, OPS v7.3.2.2

Thanks in advance

Kevin Brand Received on Sat Jun 28 1997 - 00:00:00 CDT