Adding Additional External Storage to MSSQL Failover Cluster

I have written a series of articles about Setting up and Configuring Failover Cluster with Windows Servers, with the most recent article showcasing of the use cases for a Windows Failover Cluster: Setting up Microsoft SQL Servers (MSSQL) across a Failover Cluster.

In this article, I will be writing about adding in and configuring external storage to the MSSQL Server after initial installation setup.

Pre-requisites: (Also a list if you are interested in previous articles in the Series!)

1. Setting up Failover Cluster with Windows Servers
2. Configuring iSCSI storage across Window Servers
3. Configuring Window Failover Cluster with iSCSI Shared Storage
4. Setting up Microsoft SQL Servers (MSSQL) across a Failover Cluster

Let’s start with storage options that can be mounted, other than Quorum Disk (required to ensure there are no clashes between accesses to the shared resources between the servers)!

Cluster Shared Volume as SQL Storage

Disk converted to Cluster Shared Storage is managed globally by the cluster and not an individual role. It gets mapped to a special path, typically C:\ClusterStorage\Volume1

Because all nodes in the cluster can access a Clustered Shared Storage (CSV) simultaneously at the OS level, the disk doesn’t actually need to “fail over” or move between nodes. Therefore, the cluster does not place the disk under SQL Server role (if the disk was selected during SQL Failover Cluster Installation Steps!)

Note that there are Server Name, ROles and Other Resources under the Role in the bottom tab, but there are no disks specified even when the MSSQL Server is running.

Available Storage (not added as Cluster Shared Resource) as SQL Storage

When a storage disk mounted on 1 of the servers in the failover cluster (I used iSCSI targets in my setup), it will be appear as an available storage until it had been converted to a cluster shared resource, to be accessible from all the servers in the cluster. Because only one node can physically mount the server at one time if it is added directly to the SQL Failover Cluster, it must failover.

So here, we will look at adding Available Storage to the SQL Server directly in this setup. Multiple disks can be used for separation of data routed form the SQL Server, but we will be using a single Available Storage for this simulation.

Adding Storage to MSSQL Failover Cluster

I have a Drive configured with iSCSI labelled G, detected under Storage/disks in the Failover Cluster.

Let’s go under the roles, and click on SQL Server (MSSQL Server) role. Click on Add Storage under SQL Server (MSSQLSERVER) on the right tab.

Available Storage will be detected. Click on the Cluster Disk detected, which in my case, it’s Cluster Disk 4.

This storage will be registered under Storage under the MSSQL Server role.

Under Storage/Disks, Cluster Disk 4 has been assigned to SQL Server (MSSQLSERVER) as shown in the image below.

Configuring Dependencies

Now that the available storage had been directly assigned to MSSQL Server, let’s configure dependencies to allow the storage to failover.

Under Roles/SQLServer (MSSQL SERVER) role, head to bottom tab and right click on the SQL Server under Other Resources and click on Properties.

A SQL Server Properties tab will pop up.

Under Dependencies, add in AND and select Cluster Disk 4 in the drop down. There will be a resource already registered, which is the name of the SQL Network configured previously during SQL Failover Cluster installation.

Click Apply and OK to close the tab.

Now, you can fail over across nodes in this Failover Cluster Instance setup with the storage specifically assigned for MSSQL. ♪♪ｖ(⌒ｏ⌒)ｖ♪♪