I love me some VMWare! I think it is one of those killer apps that has revolutionized Information Technology. Think about it, with virtualization you can save money on hardware, and still scale out your server environment. Also, it allows you to perform disaster recovery easier, because you can restore entire servers in a matter of hours instead of days. Also, VMWare specifically has made their product very easy to use and it is very intuitive.
That being said, setting up a Microsoft Failover cluster isn’t as straight forward as you would want. It took my boss and I a little while to figure this out on our own since our VMWare certified colleague is no longer with us to hold our hands. Still, we got it sorted out, and I thought I would share with you how we were able to do it. By the way, this guide is for Windows Server 2008 R2. There may be slight differences if you do this in other versions of Windows.
- Deploy two identical servers with only one hard drive to start. Lets call them ServerA and ServerB
- Join them to your domain, then go ahead and power them off
- Provision Two LUNs on your SAN or NAS. One with a minimum of 512KB for use as a shared Quorum disk. The other as big as you want.
- Make sure your VMWare hosts can see the LUNs.
- On ServerA, Edit the settings and add a new Hard disk using Raw Device Mappings, and select your the LUN you created for a Quorum disk
- Assign the disk to SCSI (1:0), This is important!
- Ensure that the SCSI Controller Type is configured as LSI Logic SAS (This is important too!)
- Set the SCSI Controller’s BUS Logic to Virtual if ServerA and ServerB will be hosted on the same VMWare host, or Physical if they will be on different hosts. If you have DRS enabled, I recommend using Physical.
- Do the same thing, and add the second LUN, and assign it to SCSI (1:1) on ServerA.
- Now on ServerB, we will add the two disks, but a little differently.
- Instead of adding Raw Device Mappings, you will select to use an Existing Virtual Disk
- Browse to the disks you created for ServerA (It is easier if you store them with ServerA)
- Assign each disk to SCSI (1:0) and SCSI (1:1) like you did on ServerA
- Make sure you set the same settings for the SCSI controller as you did in steps 7 and 8.
- Now the groundwork is ready, it is time to power on the VM’s
- On ServerA, Bring the new disks Online, Initialize them and Format them with NTFS
- On ServerB, just bring the disks online
- On ServerA and ServerB, Install Failover Clustering from Server Manager > Add Features
- After a reboot (If required) on ServerA Open Failover Cluster Manager From Start > Administrative Tools
- Select Create a Cluster and follow the wizard adding both ServerA and ServerB
Once all the above is complete, you are ready to configure whatever application you were planning to use Failover clustering on. SQL, IIS, whatever!
Did this help you out? Did my steps make sense? Got questions? Hit me up in the comments!