1

I'm a application manager, and I had my application wizard create a SQL database on the clustername, not on one of the two nodes of the cluster, but on the clustername.

Now a week later, my application stopped working. After some research I found out, that the SQL cluster did a failover.

I've contacted the guy who created the SQL cluster, and he told me;

-It's your own fault, you did not create the database in a correct way, you have to set some flag on the database. And you have to create a user on both nodes. (https://support.microsoft.com/en-us/help/918992/how-to-transfer-logins-and-passwords-between-instances-of-sql-server)

I was amazed by this reaction, and thought the cluster was 'transparent'. And I assumed that once a database is created on a clustername, the cluster would garantee uptime during failover, and create the users on both of the nodes of the cluster.

So the question is: On a 2016 SQL cluster that is;

  • properly configured
  • installed using the Microsoft recommendations

If you create a database on the clustername (not on one of the nodes), can you assume it's automaticly high available?, Is this database still available if a failover occurs?

Dennis
  • 677
  • 1
  • 9
  • 23
  • 1
    Is this a true cluster with shared storage? Or, are you referring to AGs? With a cluster, you connect to the sql instance using the Sql Server Network Name, slash, instance, if a named instance. For AG's, you would need the login created on both the primary and secondary servers, just as you link describes and you would use the "listener" to connect to the Sql Server. – rvsc48 Jul 05 '17 at 17:25

0 Answers0