failover in RHEL4 with Oracle

[gab_in]

Hi,

We have an application which runs on Oracle 10g with RHEL 4(RedHat). We want to configure it in our two sites in such a way that if
one site goes down for some reasons it can switch over to another site automatically.

Can somebody give us the solution for this ?

Thanks in advance

[Steven E. Protter]

Shalom,

Several.

They key issue is data replication.

You can use Oracle data guard to replicate data between the sites. Oracle streams and other options also exist.

Then you can use Red Hat Cluster Suite or HP Serviceguard for Linux to handle failover between the sites.

SEP

[gab_in]

Thanks Steven,

So I will have to first implement Red Hat Cluster Suite or HP Serviceguard for Linux for OS and hardware failover and then Oracle data gaurd for database and application failover. please correct me if I am wrong.

Waiting for your reply.

Thanks

[Delrish]

Hi,

Which Edition of 10g are you using? R1 or R2?

If you are using 10gR2, you do not need to use clusterware software.You can use Fast-Start failover option of Oracle.
Fast-Start Failover is an Oracle Data Guard 10g Release 2 feature that automatically, quickly, and reliably fails over to a designated, synchronized standby database in the event of loss of the primary database, without requiring manual intervention to execute the failover.

Alireza

[gab_in]

Hi Alireza,

Thanks for the response.

You mean to say we don't have to implement the OS failover on RedHat and Oracle data gaurd will take care of this ?

pls reply

[Serviceguard for Linux]

I do not believe that RedHat cluster Suite supports site-to-site failover.

Serviceguard for Linux has 2 multi-site failover solutions. What distance are the sites? What storage are you planning on? That will help determine which solution is the best fit.

If the distances are in the appropriate range, with Serviceguard you can use the Oracle toolkit and have direct failover rather than using dataguard.

[skt_skt]

We do have DB(oracle 9i) fail over configured via data guard with out any Service Guard. The DBs will be in a active/standby mode and the replication happening through the archivals.

it is not a hot standby configuration as there will be an expected data lagging and as also we prefer a manual fail over in DR scenario.