Re: Disaster Recovery...

As far as I can remember, you will certainly have problems reestablishing your primary machine. You can do that only by exporting or doing a full offline backup of the recovery machine and importing or restoring it on your primary server. That is the main advantage of parallel server option over replication (besides the speed).

Martin Haltmayer

Vivek & Riga Seth wrote:
>
> Let me describe the setup
> Application host, where the application is running and writing
> transactions to database. Database open with archive log turned on.
>
> Standby host. Database opened in recovery mode. A process running on
> the Standby host scans for the the archive logs on the Application host.
> As soon as it finds it, it copies the archive logs and applies them on
> Stand by host.
>
> During the disaster (Application Host Database gets corrupted) , on the
> Standby apply all the archive log files copied, open the database with
> RESETLOG option and backup the database. We loose couple of transactions
> which is OK. Application can be brought up on this host.
>
> Problem is how do we bring the original Application host (the one where
> database was corrupted) in Distater Mode. At present what we do is,
> restore the Database backup from the one taken at standby host after the
> Reset Log option, and then start this site in recovery mode. Now problem
> with this scheme is that the Disaster site is TRans Atlnatic (Europe)
> while primary is (US). rcp of Database cannot work for a 50 GB database.
> Tape transfer means turn around time of 48 hrs. This needs to be
> reduced.
>
> Any bright ideas or questions. Please send them to
> gsvs_at_mnsinc.com.
>
> Will post after consolidating all the info...
  Received on Fri Oct 31 1997 - 00:00:00 CST