SQL redundancy question

I have a 64 bit SQL 2005 server that is running our mission critical program.  I back it up every night but if something bad happened we could loose tens of thousands of dollars per hour until it's back up.

I have some money and approval from the suits to build a redundant system.  I planning on buying another server and sql 2005 and have the Database Mirroring and Failover Clustering configured to always have the system up.

My shares are in a DFS setup and it's working very well.

Is my proposal to build a whole new server the best solution?  Is there a better program to replicate the information between servers?  Can I run the SQL instance in a DFS namespace?
Who is Participating?
Two things to watch out for (out of the many):

1. A cluster typically needs to be built from the ground up as only specific hardware can be used. This might impact your existing system.

2. In a mirrored configuration the speed of the network connecting the principal and the mirror is critical to the performance of the mirror, especially in syncronous mode. I would always recommend a dedicated connection for the mirror running twice as fast as you need - which only testing will determine.

I prefer the mirrored config as it means you can have a completely seperate system in a different building or location and avoids the issue of the shared disk in a cluster. In a cluster the shared disk is still a Single point of failure.
You can go two ways:
1. build a complete Windows cluster, including a SQL Server or
2. build two separate Servers, run SQL Server on both and mirror the database (you want a third very small SQL Server instance to fulfill the witness role in the mirroring scenario)

But you need the second machine in both scenarios.

Hope this helps ...
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.