Standby db or replication

Hi,

We are required to setup an estimated 2 Tera byte Db for an online transaction processing project.
The db has to keep up a high level of service level. (24x7) Besides there are also large regular back-end processing jobs.

The problem is that there are also requirements to do a substantial amount of reporting from the data on this database. Possibly some large backend and also user reports.

My peers are suggesting we do replication on the database to a seperate machine for the reporting.
My feeling is that doing replication on 2 tera of data is quite unthinkable and I have no experience on what the performance impact on the db wiil be like.
(It could be better off to let reporting be done on the db itself. We are talking about possibly 3hourly snapshots).

I was also thinking of setting up a standby database instead and opening it as only read-only database.
Though with only limited availability, it serves its purpose with less overhead.
I have read somewhere that 8i can do that but the current version we are on is 8.0.5 so I'm not sure it can be archieved. I know that standby databases, once opened, will no longer be in sync with the master
db in earlier versions.

Anyone has such setup or experience and care to share how you have overcome the problem ?
I'm open to any suggestions and comments

Thanks Received on Wed Jul 07 1999 - 20:25:31 CDT