Steps to Set up HPC Pack with Failover Clustering for the Head Node
Applies To: Microsoft HPC Pack 2012, Microsoft HPC Pack 2012 R2
For a list of important requirements for setting up Windows HPC Server 2008 R2 in a failover cluster, see Requirements for HPC Pack in Failover Clusters.
To set up HPC Pack with failover clustering for the head node, you need to set up SQL Server to support the head node, create a failover cluster, and install the head node services in the failover cluster. The following topics provide details about this process.
Note
This document describes how to set up a remote installation of SQL Server to support the head node in a failover cluster, which is often a recommended configuration. You can use a local installation of SQL Server instead, that is, where your failover cluster runs both SQL Server and HPC Pack. To create this configuration, follow the steps in the list below, but when you come to "Step 5" (for setting up remote SQL Server for the HPC cluster), install and configure SQL Server for high availability on the same servers that you will use for the head node failover cluster. Then, when you reach "Step 6" (for installing and configuring HPC Pack), do not specify a remote SQL Server instance, but instead specify your local SQL Server clustered instance.
Step 1: Install Windows Server on Multiple Servers
Step 2: Set Up Shared Storage for Servers for the Head Node
Step 3: Set Up Failover Clustering and File Services for Servers that Will Run the Head Node
Step 4: Configure a Clustered File Server for the HPC Pack Head Node
Step 5: Set Up Remote SQL Server for the HPC Cluster
Step 6: Install HPC Pack on a Server that Will Run Head Node Services
Step 7: Configure the Head Node on the First Server
Step 8: Install and Configure HPC Pack on Another Server That Will Run Head Node Services
Step 9: Validate Installation of Head Node Services in the Context of the Failover Cluster