0

I need 2 SQL Failover clusters:

  • Core Database Cluster: this contains all of my core application databases
  • Analytic Database Cluster: this contains my web analytic database

Now, I only have 2 physical servers (DB1 and DB2) connected to a SAN. How to deploy two SQL Failover Clusters in my current database servers?

I assume that DB1 is set as the primary node for my Core Database Cluster and secondary node for the Analytics Database Cluster. And DB2 is set as the primary node for my Analytics Database Cluster and secondary node for the Core Database Cluster.

Is this the right way?

slm
  • 7,355
  • 16
  • 54
  • 72
ronanray
  • 103
  • 1
  • 3

1 Answers1

0

Just set up two clustered instances in ONE cluster. Instance1 on DB1 and Instance2 on DB2. Give each instance it's own shared disk(s).

Instance1 can reside on either node, but the preferred node would be DB1 Instance2 can reside on either node, but the preferred node would be DB2

Mark Allison
  • 2,098
  • 7
  • 26
  • 45
  • Right. So, for the SAN storage setup can I assume that EACH SQL instance would require: 1 x LUN => SQL Data Disk; 1 x LUN => SQL Log Disk; 1 x LUN => MSDTC; Any recommendation to achieve best performance for this kind of setup? Thx – ronanray Aug 31 '13 at 09:29
  • @ronanray The minimum required is 1 shared disk per instance, but you go with what you need. If using more than one disk per instance, I usually set up mount points for the drives. That way you get one drive letter per instance (as you can quickly run out of them). MSDTC is not required on SQL2008 and later I believe, although I could be wrong. – Mark Allison Aug 31 '13 at 12:36