Virtualization: Hardware Consolidation
Done properly, consolidating server hardware through virtualization can convey many more business benefits than simply reducing hardware costs.
Adapted from “Microsoft Virtualization” (Syngress, an imprint of Elsevier)
The traditional mindset around consolidation has always been to simply reduce the number of servers in a datacenter. Since then, we’ve realized how consolidation is much more than simply reducing your hardware footprint.
Done properly, you should go through workload consolidation and hardware optimization before you start thinking about true hardware consolidation. This discussion will be based on the assumption that you’ve already identified these items and considered them for your new dynamic datacenter.
Sometimes it may be justifiable to pay for extended support of end-of-life (EOL) hardware, rather than replacing it with new equipment. This may be a hard concept to understand, but some will get it right away. You simply need to balance your expectations, costs and technology to see what combination fits best.
Support or Supplant?
Let’s consider as an example a low-impact, development-only server virtualization farm. Here’s the concept: Your development staff needs to create a farm of virtual servers they can use for testing new applications. This farm won’t support production users. It won’t be expected to be available and online 24 hours a day, seven days a week. Meanwhile, you’ve recently retired a complete blade server chassis due to it reaching EOL.
You’ve been diligent crunching the numbers, and management realizes that the cost of extended support on this blade chassis is far more than the company is willing to pay for limited support. The plan for this used blade chassis is to wipe the drives and pay somebody to haul it away and dispose of it. This is a fairly common scenario. It’s just this type of situation where you can jump in and provide your company with some additional workspace for little to no additional cost.
As long as you do workload consolidation and hardware optimization properly, it’s almost inevitable that you’ll end up with surplus hardware that has reached its EOL support. That’s typically what ends up on the list of items to be thrown away.
However, new technology and new processes can help you continue using this hardware to provide a true benefit to your company. So, how do you use an outdated blade chassis and its servers to provide a reliable development platform without paying for extended hardware support? There are a couple of different approaches. You’re starting with a 14-blade chassis complete with servers. You need a multiserver development platform. Here are two possible consolidation approaches.
Approach No. 1: Clustering
In this approach, you’re going to take the blade chassis and cluster the server environments across multiple hardware platforms. Build four server OSes across 12 blade servers. Each OS will be clustered across three server blades, reserving two of the blades to be left offline and labeled as “spares.”
This approach lets you keep running the OS, even if one of its three servers should fail. In this case, one of the two remaining spare blades can replace the failed blade while it’s either repaired or replaced. This concept is nothing new.
Microsoft clustering improved greatly with Windows Server 2003. A number of third-party vendors have also been providing clustering solutions for years. Clustering provides the needed workspace while reducing the risk of hardware failure by spreading the risk of a single OS across multiple hardware platforms.
Approach No. 2: Clustering with Virtualization
Our second theoretical approach builds on the first by adding an additional layer of protection that spans the entire chassis, rather than just the three hardware blades used in the first approach. Some may consider this overkill, but it’s worthwhile if you have the available resources.
If the hardware is extremely old, and the environment in which you’re planning to use it is extremely important, this approach may quickly change from overkill to sufficient. An added layer of protection can also increase the number of servers from four to eight. Using virtualization, you can easily back up the OS using snapshots and migrate from one cluster to another almost seamlessly.
Clustering with virtualization can work well in an environment that requires additional workspace without the desire to continue paying for extended hardware support. Depending on the technology you choose, you may have to consider additional licensing costs. Even those additional costs may be feasible if it means the continued use of hardware that might otherwise be thrown away.
Above all else, remember that hardware consolidation doesn’t have to be complex. Whether you choose to build a clustered virtualization farm out of an old blade chassis, or simply provide your development engineers outdated hardware server platforms that provide higher performance than their current desktop platforms, you are consolidating. Sometimes the simplest ideas can be the most fundamental and show the greatest business benefit.
Thomas Olzak is the director of information security at HCR ManorCare, an Ohio-based short- and long-term rehabilitation and medical care provider with more than 500 locations spread throughout 32 states. Jason Boomer*,*Robert Keefer and James Sabovik also contributed to this article and the book from which it’s excerpted.
©2011 Elsevier Inc. All rights reserved. Printed with permission from Syngress, an imprint of Elsevier. Copyright 2011. “Microsoft Virtualization” by Thomas Olzak. For more information on this title and other similar books, please visit elsevierdirect.com.