Building a lab datacenter from the bottom up

In the coming weeks I am going to be making several posts, videos, and screenshots about building a lab datacenter.  The why’s and how’s to this are pretty lengthy, however I will be taking a short answer to the why’s because basically it is for education.  Your goal in this field, if you want to exceed (rather than just succeed) then you must have a desire for education and knowledge.  Most of us learn the best by doing.  In the coming labs we will be doing a lot of doing from setting up storage, installing Vmware ESX , installing Windows 2008 R2 (several times), installing Windows 2008 SQL Server (for vCenter, and later for Vmware view), configuring DNS and maybe some other things depending on what comes up during the process.

It is expected that the lab will change in that we get some new hardware to incorporate into the lab, we’ll make the necessary adjustments and push forward.  Along the way we’ll be taking a look at the importance of planning your environment to avoid some potential “gotchas” as we go along.

One of the first things to think about as far as that is concerned as I plan to leverage a new feature in Vmware ESXi 5 known as Auto Deploy.  This will allow you to boot servers without local storage.  Today this is known as PXE booting, roots are from BOOTP and it all basically still works the same.  However, what happens in a total outage?  Your Auto Deploy environment needs to have DHCP, DNS, vCenter all available in order to operate itself.  Therefore we will create a architect around this issue by creating a Tier-1 environment with local storage to avoid this issue.