Microsoft made several improvements to its hypervisor with its Windows Server 2012 release, but understanding new features and getting started with Hyper-V can be intimidating. In a previous interview, I talked with author and Microsoft MVP Aidan Finn about how he thinks these new features are helping Hyper-V change the virtualization market. In this Q&A, Finn discusses how an organization should approach a Hyper-V deployment, including...
planning considerations, common stumbling blocks and learning resources. Not surprisingly, one of those resources is the new book he co-authored -- along with Patrick Lownds, Michel Luescher and Damian Flynn -- called Windows Server 2012 Hyper-V Installation and Configuration Guide.
The book is available now and you can download a free chapter except (PDF). In Chapter 2, the authors lay the groundwork for deploying the new operating system, including points to consider when getting started with Hyper-V. Chapter 2 also covers the installation process, reviews new features and explains how to export and import virtual machines. This chapter excerpt is available courtesy of the publisher Sybex, a part of John Wiley and Sons Inc.
What basic steps should an organization take before getting started with Hyper-V?
Aidan Finn: I am a huge believer in performing some kind of assessment. I believe that correct sizing of the environment must be done scientifically. This applies no matter what virtualization platform you choose. An assessment allows you to identify virtualization candidates, measure their static requirements (such as disk space) and record their low, average and peak resource requirements (such as CPU and memory) over a length of time. This operation is relatively simple and Microsoft provides the free Microsoft Assessment and Planning Toolkit. The resulting reports give an insight into the environment and allow you to scientifically determine the sizing and licensing of your new virtualization environment, including host hardware, networking and storage.
Those who have worked in the hosting business, like myself, know that cloud computing makes the assessment a bit pointless. The point of the cloud is that the administrator has enabled self-service so they never know what is coming down the pipe. For example, IT deploys a private cloud, and the various departments or divisions consume resources and pay for what they use. You can't size that.
A proof of concept is incredibly useful. This allows the engineers or consultants to identify the best configuration of the compute cluster [the hosts] and fabrics [networks and storage] before the system goes into production. The environment can be tested and developed with production-like systems. The usage and costs of the resources can be figured out, and this data can be used to determine an algorithm for cross charging the cloud customers, enabling IT to become a nonprofit/loss making operation. Alternatively, it might be useful for forecasting future costs where the business sees IT as a loss making cost-center. And, of course, this allows the engineers, administrators and operators to become more familiar with Hyper-V and System Center and create their processes and runbooks so that things go more smoothly in production.
What area of deploying or configuring Hyper-V do administrators often have the most trouble with?
Finn: Two items have traditionally caused difficulties for those getting started with Hyper-V:
- Backup of Cluster Shared Volumes (Hyper-V's alternative to vSphere's Virtual Machine File System)
Windows Server 2012 gives us a more mature version of Cluster Shared Volumes. The biggest challenge was how this storage was backed up when we were protecting our clusters of Windows Server 2008 R2 Hyper-V hosts. The under-the-covers process was very complicated and, ideally, we should have been planning storage design and VM [virtual machines] placement around backup plans. That was like putting the horse before the cart for most people. Windows Server 2012 Hyper-V gives us a greatly simplified backup mechanism with a single reliable and distributed Volume Shadow Copy Service snapshot for application-consistent backup of VMs.
We know that networking has proven to be a challenge for some. There are many reasons. Some early Hyper-V adopters were new to virtualization so there was a learning curve for them. Not many people had ever created a Windows Server failover cluster so those networking requirements were new. We also saw people coming over from VMware who assumed that Hyper-V worked the same way; Hyper-V and vSphere are both Type 1 hypervisors and many concepts are similar, but how VMware and Microsoft do things can sometimes be very different.
Networking has grown a huge amount in Windows Server 2012. There is an incredible amount of new features in this release; I wrote nearly 100 pages on the basic topics for our new Hyper-V book -- and that was before we got to the advanced cloud networking features. In the past, there were only a few ways to do networking in Hyper-V. Now there are an amazing number of options. Maybe the best option is to do something like a car racing team. Do the proof of concept, and try one new thing at a time. Repeat the tweaks to learn the technologies and identify the best solution for you. I think some consultants are going to settle on a small number of variations of hardware and configuration. They'll build up great expertise in these few recipes and give their customers stable and predictable results.
What resources would you recommend to people getting started with Hyper-V?
Finn: I might be biased, but I hear there's a new book published by Wiley called Windows Server 2012 Hyper-V Installation and Configuration Guide. That was a project by myself, Patrick Lownds (MVP), Michel Luescher (Microsoft Consulting Services), Damian Flynn (MVP) and reviewed by Hans Vredevoort (MVP). We put over six months of work into learning this product and documenting it, hoping to give the reader more information than they might find anywhere else.
For questions, the Hyper-V TechNet forum is a good start. There are also a number of blogs where you'll find lots of information:
- My own blog
- The Hyper-V.nu blog run by a number of experts in the Netherlands, including Hans Vredevoort
- I recommend Damian Flynn's blog. He is the best System Center person I know.
- You can find quite a bit of information about storage and networking at Didier Van Hoye's blog.
- Thomas Maurer in Switzerland is also writing a lot of good stuff spanning Hyper-V and System Center at his blog.
- The Microsoft Partner & Customer Solutions Blog has been summarizing many MVP blog posts every week.
Download Chapter 2
Download and read Chapter 2 of Finn's book, Windows Server 2012 Hyper-V Installation and Configuration Guide.
Two essential RSS feeds to subscribe to are the feeds that list the released hotfixes for:
Microsoft has made a significant investment in the Microsoft Virtual Academy. There you will find learning materials on Hyper-V and System Center, including training for those switching over from VMware.
You might also want to learn or evaluate these technologies in a lab without making a financial commitment. Microsoft makes that possible with some free time-limited evaluations.
About the author:
Aidan Finn, a Microsoft Valuable Professional (MVP) with the Virtual Machine expertise, has been working in IT since 1996, specializing in server/desktop management and virtualization. Currently, Finn is working as a technical sales lead for a distributor in Dublin, Ireland, specializing in Microsoft technology. He blogs on http://www.aidanfinn.com, tweets as @joe_elway and regularly presents on topics such as Hyper-V, Windows Server/desktop and System Center. Finn has also contributed to or written books, including Windows Server 2012 Hyper-V Installation And Configuration Guide, Microsoft Private Cloud Computing and Mastering Windows 7 Deployment.