2

I'm setting up a simple 2-node failover cluster with SQL 2014 on Server 2012 R2 in Core mode (no GUI). I have the cluster configured with identical machines, and I have the first node configured and working.

When I go to add the second node however, the wrong configuration information is used. Specifically, the shared and instance binary locations point to the OS drive (I installed them on a different drive on node 1). I tried manually specifying the locations in the configuration file, but setup won't accept these parameters when the install mode is AddNode. Did I do something wrong, or is this a bug?

2

It appears to be a bug/feature. It was discussed previously on SQL Server Central http://www.sqlservercentral.com/Forums/Topic1578488-2799-1.aspx where Ed Vassie said:

Put it another way... On an ADDNODE install you cannot specify the location for the SQL binaries, they always go to the system drive.

Ed Vassie is a reliable source as the original developer of FineBuild, a tool specifically for rolling out repeatable installs of SQL Server including in clustering mode. He lives and breathes this part of SQL Server.

1

A workaround for this problem:

Uninstalled everything, ran setup with /ACTION=PrepareFailoverCluster on both nodes first (which allows you to choose the binary locations), and then ran it again with /ACTION=CompleteFailoverCluster on one of the nodes. It takes three runs of the installer instead of two, but it solves the issue.

The only limitation is that you have to be setting up a new cluster and know how many nodes you need ahead of time. Prepare cannot be run after Complete has run.

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.