The Virtualization Room

A SearchServerVirtualization.com and SearchVMware.com blog

» VIEW ALL POSTS Apr 27 2009   2:52PM GMT

Consolidation ratios are yesterday’s news



Posted by: Craig Hatmaker
Tags:
hypervisor
Virtualization
Why choose server virtualization?

Ever notice that when you talk to someone about virtualizating, it always seems to come down to, “So, what’s your consolidation ratio?” Everyone seems to care only about the number of virtual machines you can house on a single host system. While consolidation ratios are important, they’re yesterday’s news!

Virtualization is about so much more than just shrinking the footprint (physical and carbon) of your data center. Think about it: What does virtualization really do for you? It encapsulates your workloads (servers) into a collection of files that are consistent, that are portable, that are uncoupled from hardware, and that can be copied from location to location. Let’s look at each of these benefits individually:

Consistency: A VM is simply a software representation by a hypervisor to a guest operating system of the interfaces required to interact with the “core four” compute resources (CPU, memory, disk and network).
Portability: A VM that you create on one virtualization host can be moved to another virtualization host with impunity. Today, you’re restricted to a single hypervisor solution. As standards develop, I believe you will be able to move VMs among different vendors’ hypervisors without modification — either because everyone has converged on a single standard, or because there are few enough standards that competing vendors can implement emulation or translation services to enable VMs to “float” from hypervisor to hypervisor (cloud pun intended).

Uncoupled: Since these interfaces are implemented in software, they can be abstracted from the underlying physical system and offered to the guest OS in a very consistent manner. There are no more concerns about whether you have an Intel or a Broadcom network adapter, or an LSILogic or a QLogic HBA — or even local disks or centralized storage served up from a SAN or NAS. In fact, it doesn’t really matter if the underlying hardware is from Dell, HP, IBM, Fujitsu Siemens or is a white box solution … the VM and the guest operating system within it don’t know and don’t care.

Can be copied: This is an interesting one. Many people consider this to be a risk of virtualization. Not only is it possible to capture the data used by an application, but you can copy the underlying files that comprise the virtual machine and have the entire operating environment. This poses significant security concerns, as well as licensing concerns. Since a VM can be copied, you’ve essentially made it possible to put your servers onto a USB or other removable media device and walk away with them. All true, all valid concerns (that’s where your security policies come into play). But look at the benefits you derive from being able to copy entire VMs! You can now perform stateful backups of an entire server. No longer do you have to worry about whether your system state matches your application’s data state — they’re in perfect sync. If you have to restore a server due to failure or “operator malfunction,” you simply copy the files that comprise the VM from your archives and you’re back up and running. No more installing the base operating system, applying system state, installing an application and restoring data (then praying that everything works). It’s all rolled into one simple, seamless process.

As you can see, there’s a lot of benefit there with nary a mention of consolidation ratios. But how can you take advantage of those benefits? One of the most obvious ways is to enable rapid provisioning of new servers. Think about your current physical environment. How long does it take from the point where you get authorization to purchase, until you have a server sitting on the data center floor configured with your standard suite of software (operating system, applications, tools, etc.)? In many enterprise organizations, it can easily take six weeks or more! In certain verticals, it can take even longer, due to certification/validation requirements.

In a virtual environment, from the point of getting the signature, you can have your new server deployed and configured in a matter of minutes. And that’s not just any server, but a server that has already gone through your internal validation procedures and been blessed to connect to your network! How much value would your business derive from being able to respond to a change in customer needs six weeks more quickly than you do today?

Another area that’s ripe for improvement in most organizations is disaster recovery planning and implementation. Look at your current DR plan (if you have one and you can find it). What is your current recovery time objective (RTO) and recovery point objective (RPO)? If you’re like most organizations, your RTO for critical services is on the order of 48 hours. Being generous, your RPO is something like 24 hours. When you add the two together (assuming you actually hit your RTO), you get the length of time for which there will be NO information available about your business. How much downtime/data loss can your business afford?

Since a VM is nothing more than a data file (as far as the hypervisor is concerned), it can be replicated to a remote site using technologies that have been around for years! Now you can have a DR plan that you can test at will, that embraces the use of “legacy” systems  — no requirement to have “like” systems for production and DR! –  and that includes RTO and RPO values that make your old DR plan blush.

On another topic, when was the last time you did a platform upgrade for one of your business critical systems? How long did it take? You had to plan it, test it, certify it on the new platform, schedule enough downtime to effect the migration, test it again and finally went off the air and did the upgrade … over the long holiday weekend … to give yourself that extra buffer … just in case. How many change control requests was that? How many man-hours invested? If your systems were virtualized, the upgrade could have been tested, verified, scheduled and executed all within a couple days. It could be done with zero downtime, although you might want to take the opportunity to grab a fully quiesced backup of the system in a powered-off state (something that’s always good to have). And, oh yeah, you could have been home with the family enjoying that long weekend!

So you see, there’s a lot more to virtualization than just consolidation ratios. Virtualization is the foundation for building a robust, flexible, responsive business enablement center.

1  Comment on this Post

 
There was an error processing your information. Please try again later.
Thanks. We'll let you know when a new response is added.
Send me notifications when other members comment.

REGISTER or login:

Forgot Password?
By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy
  • KenLC
    All the points are valid. However, compression ratios are probably what is going to justify the move to virtualization in the first place. If you can't get the project approved, it is hard to get to all of those other nice thngs.
    60 pointsBadges:
    report

Forgot Password

No problem! Submit your e-mail address below. We'll send you an e-mail containing your password.

Your password has been sent to: