<h1>How to use the Virtualization Lab (I)</h1>

I finished last post on this series with a fully working cluster installed between two Hyper-V virtual machines (VM) using a virtual iSCSI solution installed on a Virtual Box VM as depicted in the next picture:

 
Before moving on in the process of adding complexity to the lab scenario, don't forget to safeguard your work; although this just a lab, it doesn't reduce the nuisance of having to reinstall everything in the event of any failure. So, create VM snapshots:




And setup and execute some 2008 SQL Server backup options:





In order to prepare my setup for the next steps I installed a second NIC on my desktop PC, to overcome the limitations of trying to use two simultaneous bridge connections over a single physical NIC, and expanded the size of the iSCSI target used by SQL Server just to be on the safe side before installing databases on it.



The iSCSI target expansion can be made without stopping the service.



I double the size of the disk used by the clustered SQL 2008 to 10 Gb



The newly added space now appears has unallocated space that can be used to expand the already existing partition. All of these expanding operations were executed with both the cluster and SQL Server running.



As promised, I kept on improving this setup and the next step was the installation of the System Center Virtual Machine Manager 2012 Release Candidate on another VM. Yes, you guessed it, on a VMware VM!

When done, I had this:



Although this is a very simplified picture, you can see I had SCVMM running on a VMware VM with databases installed over the SQL server running on the Hyper-V cluster, which is using the iSCSI storage from the Virtual Box VM.

In the next picture you can see both VMware and Virtual Box running with SCVMM 2012 RC correctly displaying the VMs running on the Host-Server where I installed the  required Agent.



But that was not the end of it!

Next, with the cluster running, I moved both services (MSCDT and SQL Server) to Node 1, evicted Node 2 and added Node 3 to the cluster following the same basic procedures previously explained.
Once done with that, I installed System Center Operations Manager 2007 R2 on a new Virtual Box VM also using the clustered SQL server as the database repository. The first step was the installation of the database on one of the nodes (the one owning the SQL server) and then the manager itself on the LAB-SCOM2007 VM.



This way I managed to have a two node cluster, one on Hyper-V the other on VMware, both using iSCSI storage from Virtual Box and being managed by SCOM 2007 also in Virtual Box.



In any of these three designs one has plenty of room to test and learn how to use and abuse from all the features available in Windows 2008 Failover Clustering.

Keep reading, there is plenty more to come…