Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: How to replicate a Database?

Re: How to replicate a Database?

From: <nasof_at_hotmail.com>
Date: Fri, 27 Nov 1998 06:28:37 GMT
Message-ID: <73lgqe$cnu$1@nnrp1.dejanews.com>


If you set up Oracle's Advanced replication for your tables (see the manuals for how to do this...) I would recommend using Full-time Software's "Octopus" for failover. They have proven software on UNIX and NT, and were recently purchased by Legato systems (the tape Media Layer software that ships with ORacle for NT. You can find out more info at: www.qualix.com

Basically, you create a database with Replication Support and adv. replication as you want it (determining conflict resolution is the toughest part) through the Replication Manager GUI interface. Set it up so that both nodes are masters (ie. both can have data entered into them.) Next you install Octopus (about $1000-1500 per node) which will basically switch IP addresses during the failure of another node. This way, you dont have to recreate the primary database each time you fail over (as in stand-by.)

In article <01be195e$78fe3b20$2bcbfcc1_at_moi>,   "Damien" <spanish.inquisition_at_altavista.net> wrote:
> > We are running a mission critical Oracle database. We have 2 dedicated
> > serves running Oracle 8.04 on NT4.0. How can we setup the primary Oracle
> > server to replicate all its data to the backup Oracle server. We would
> like
> > to have hot standby. Does Oracle offer some tools to do this and how much
> do
> > they cost. I've read about Oracle FailSave but this requires MS Wolfpack
> and
> > NT 4.0 Enterprise which we don't have.
>
> On ora734/Solaris, I used the standby mechanism. It relies on the
> archivelog feature.
> Archivelogs are transferred (e.g. with rcp on Unix) and replayed to the 2nd
> server.
> The 2nd database runs in standby mode and it only has the capability to
> replay archivelogs. It can't be accessed for queries.
>
> Shortly, setting a standby DB consists in :
> * shutting down the 1st DB;
> * copying the tablespaces from the 1st to the 2nd server (e.g. with rcp on
> Unix) ;
> * generating "standby control files" on the 1st server;
> * copying these files on the 2nd server;
> * activate archivelog mode ont the 1st DB;
> * start the 2nd DB in standy mode (it's not fully mounted).
> * start normally the 1st DB;
> * activate a mechanism of your choice for transferring the archivelog files
> and
> having the 2nd DB replay them.
>
> Refer to your Oracle documentation for details and syntax.
>
> Activating a standby DB so it can be normally accessed (when the 1st DB
> crashes)
> is NOT reversible. Once it has been mounted, you have to repeat the full
> process above
> to re-create a standby architecture.
> Enjoy !
> Damien.
>
>

-----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own Received on Fri Nov 27 1998 - 00:28:37 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US