I understand that to achieve some level of availability of an online application it is possible to organize an environment with a redundancy using an IP failover mechanism and switching between several identical servers in the event a problem arises in any one of them.

But what about with databases? Should the database be on each server with some sort of replication data between them or should it sit on its own dedicated server?

What approach usually is used in SQL Server 2008 R2 or SQL Server 2008 R2 Express?


You'll want to put the SQL Server on its own server. You may also need to protect the SQL Server from reboots or failed hardware by looking at an HA solution. You can use Clustering (which requires shared storage), or database mirroring (with or without a witness), or peer to peer replication.

  • 1
    Generally I would say use clustering within a datacentre and mirroring between datacentres. We have several systems like this catering for a range of failure scenarios. – Gaius Jan 11 '11 at 13:42
  • If a SAN is available then Clustering is an option. If there is no shared storage then clustering is out and mirroring and replication become the options. I would normally use clustering within the data center as well, but I don't know @rem's config or budget. – mrdenny Jan 11 '11 at 21:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.