I wanted to run my thoughts by the gurus here. I think I'm on the right
track, but I just want to get some confirmation.

This morning I tried to upgrade our NT4(SP5) SQL7(SP2) 2-node cluster. When
I ran the Failover Cluster Wizard on the primary node and broke the cluster,
it apparently did not automatically remove the SQL IP and SQL Name shared
resources. This caused two issues: the SP4 install attempt failed (I
received an error: "An active SQL cluster has been detected. Run the SQL
Server Clustering Wizard on both nodes to disable the cluster before
installing"); when I tried to re-cluster later, it complained that the IP
address already existed. Does anyone know why these resources weren't
auto-deleted? Is it a fair assumption that if next time I run the FCW and
break the cluster, and I then go into Cluster Administrator and remove the
two shared resources, that I will avoid the issues I encountered?

Thank you for your time.

I cannot tell you that you will not run into the problem if you manually
remove the entries in the cluster administrator. However, after the
failover wizard unclusters, if there are entries in the cluster
administrator related to SQL Server, you can manually remove them.

Rand
This posting is provided "as is" with no warranties and confers no rights.

I guess what I'm asking is, how does the SP determine that an active SQL
cluster is running? Is it by the presence of those two shared resources or
an active process or a running service, or.... ?

Thank you

"Rand Boyd [MSFT]" <rboyd (AT) onlinemicrosoft (DOT) com> wrote

Turns out that what I went through before "fixed" the issue of the SQL
Failover Wizard not removing the old shared resources. It now does, and
everything worked out fine this morning!

Cheers

"Kripes" <kripes (AT) nospam (DOT) com> wrote