Customer Reviews

Blade Servers and Virtualization: Transforming Enterprise Computing While Cutting Costs

5 star:		(5)
4 star:		(1)
3 star:		(1)
2 star:		(0)
1 star:		(1)

 

 

It doesn't take very long to figure out that the authors really like blade servers.

They like their space savings, cable plant simplicity, and uniformity. There are very detailed (although never critical) descriptions of each manufacturer's product lines and some very exhaustive matrices comparing not only blade servers but virtualization products as well.

My problems are in the details. First of all, it is clear that in many cases the benefits being credited to blade servers accrue ONLY because those blade servers are running virtualization products. Moreover, in many cases it is unclear why the benefits would not have accrued to commodity scale out servers running virtualization products.

Two, there are some claims that just stretch credulity too far. For instance, it is claimed that a provisioning a new blade server can be done in a few hours while doing the same with a rack mount server can take weeks.

Bearing in mind that Google, Yahoo, and Microsoft have all constructed their massive data centers with rack mount servers, you have to wonder how they could have made the same mistake (from the author's point of view) several hundred thousand times for each vendor.

The fact of the matter is that blades are a perfect choice for certain data center environments, but not others. Today they constitute ~ 15 % of all server shipments and most vendors hope to get that up to ~20% by 2010.

What's the hold up? Well, blade servers cost more per unit of computing power. Their form factors are proprietary, so their is considerable vendor lock-in (and thus higher prices). And the switches built for each blade center are vastly more expensive and slower to adopt new features than standalone modular or fixed form factor switches.

Conversely, virtualization is a wonderful add-on for certain types of compute applications, but not all. (For the technically inclined, if compute latency is at a premium, or the computations requires large amounts of state to be held in memory, then virtualization is going to be at best a mixed blessing and quite possibly make things worse.) Such limitations are alluded to peripherally, but never spoken to head on.

That's a darn shame. Had the authors included a chapter on when blades make sense and when virtualization makes sense, this would have easily been a 5 star book for me. The writing style is lucid and flows well. The authors are clearly experienced in their field. However, the assertion that blades and virtualization are a broad based panacea for most types of organizations and data centers just does not conform to the facts.

We knew we were running out of space and power in our data center, and a blade environment might be the solution. But how do you even begin amassing the information to make the right decisions to solve our dilemma? I consider this book necessary reading for anyone considering the plunge into blade technology. Not only did it address my questions and concerns, it introduced me to potential problems and solutions I hadn't even considered! Now I feel confident that I have a valuable tool to aid me throughout the design and implementation stages of our conversion from stand-alone servers running at 5% utilization to blade chassis running multiple virtual servers at 50% utilization. The authors clearly know and address the benefits and pitfalls of this relatively new technology. I especially enjoyed the case studies. They allowed me to see what difficulties other companies in similar situations faced, and how they solved their problems. If there is another book out there that details these issues in a vendor non-specific format, I haven't found it.

This book is a great complement to the multitude of virtualization books out there (my own included). Hardware design and sizing for virtualization is critical in all virtualization migration projects and this book does a great job examining all of the available blade solutions and architectures at your disposal. The authors also provide granular analysis of all major vendor solutions (HP, IBM, Dell, Egenera, etc.). Coverage of clustering and virtualization software is also included.

The scope of the book is vast (blades, virtualization software, power and cooling, management -- to name a few topics), so it's unfair to expect every detail of every topic to emerge in the text (the book would be 10,000 pages if it did). But if you want a book that connects all of the dots in the modern data center, then this book is for you. With coverage of all data center technologies and the vendors that provide them, this book is sure to accelerate the preparation and architectural work needed in any virtualization migration project.

~Chris Wolf

I found the book to be an excellent source of information on various virtualization topics - hardware, software and vendors. It provides introductory information for those trying to better understand the concepts of virtualization and blade hardware. In addition, there are comprehensive chapters - Assessing Your Needs, System Software Considerations and Differentiating & Selecting, all of which are useful for those trying to evaluate their virtualization options.
I have recommended individual chapters to others such as the chapters on Benefits of Blades & Virtualization and Power & Cooling.
Finally, the case studies are very useful in assessing how other companies (perhaps like yours) have deployed virtualization.
Excellent as a reference source.

Incredibly simplistic and obvious. Approximately 1/4 related to specific product information guaranteed to be out of date. Useless.

Blade Servers and Virtualization: Transforming Enterprise Computing While Cutting Costs is a great read on a complex subject. Virtualization is taking the information technology world by storm. For investors, the question is where to invest besides the pending/probable VMware IPO? This book has answers! And they are in clear english prose. No technobabble, no sidetracks down dark ratholes.

The first 2/3 of the book is an excellent technology bible. It clearly explains the why and how of virtualization technology and relationships with it and other technologies. The last 1/3 presents case studies and examines a wide range of suppliers. This was particularly valuable as the case studies explained the economics of Virtualization and why customers are willing to spend money on it. Investor ought consider starting with the case studies, shifting to the technology and finishing with the vendors.

Not much to say except that the version of this book available as of Jan 2011 was published in 2007. It's pretty out of date, at least in terms of the information it provides about specific vendors products.

Other than that, it's a great introduction to the subject matter. It won't make you an expert, but it will give you some direction if you're just starting out with the technologies covered.

Very well-written book on supposedly complex subject matter. Gets to the point/s quickly. Leaves you feeling enlightened and eager to plan and implement rather than confused with more questions than answers.

I wish all of the IT books and literature I have read were this straightforward and productive. This is a must read if you are in the early stages of planning and implementing blades and/or virtualization in your data center.