How will VMware compete in an increasingly crowded virtualization space? On virtual architecture maturity and stability, says VMware senior director of product marketing Bogomil Balkansky. SearchVMware.com caught up with Balkansky to talk about VMware's pricing strategy and how their flagship product, ESX, stacks up against Hyper-V and others in terms of value.
What is VMware's pricing strategy moving forward in light of Hyper-V?
Balkansky: I don't think that our pricing strategy and the principles of how we do our pricing is going to change. Obviously, we always take a critical look at how we package and price our products. The objective is to make sure that we have the right products at the right prices for all types of customers. I don't think that principle is going to change.
I would argue that right now, we have the right product packages with prices that allow the customer to do whatever it is they need to do with our technology. Whether it's something simple like partitioning a server for test and development, a single server consolidation, production server consolidation for a larger environment or if they want to reap the other benefits of virtualization, such as business continuity, disaster recovery and dynamic resource management, they can do that at a price that affords a tremendous ROI. Virtualization is one of those technologies where the ROI is quick, undeniable and easy to see.
do you respond to people who are quick to point out Hyper-V's price in relation to VMware ESX's price?
Balkansky: I would say that VMware absolutely has the better product. There are multiple things that you need to take into consideration when you compare the two products. Some people make uneducated and unfair comparisons. Hyper-V is a first generation hypervisor and nothing more. VMware provides a platform that is much richer than just a hypervisor.
For a fair hypervisor to hypervisor comparison, you would have to compare Hyper-V to ESX. ESX is just a portion of our VMware platform. It's an important part, but just a part nonetheless. Hyper-V is a first generation, unproven product bundled with the operating system. You can buy ESXi off of our website for $495 or you can get it embedded in server systems from a range of vendors.
Dell made an announcement around their strategy and the fact that they offer our hypervisor on their entire PowerEdge line; HP is offering ESXi on their highest volume server models; Fujitsu Siemens has been offering ESXi since the end of February; IBM is offering it on two models. So as customers are getting Hyper-V as part of their OS purchase, they are also getting our hypervisor as part of their hardware purchase.
So, we've established a kind of parity in order to let the best product win. We've established a level playing field with Microsoft in terms of marketing power and ability to get our respective hypervisors in the hands of our customers. And we are very comfortable and confident that we have the better product if customers have a chance to experience both.
What makes ESX the better product?
Balkansky: Foremost is maturity and stability. Our hypervisor has been on the market and stress tested for 8 years in tens of thousands of customer deployments in critical applications. I have a slide showing ESX running on 4 years of uptime. You show me a screenshot of a Windows copy being up for 4 years and I'll rest my case. In short when it comes to stability and maturity of the platform, we have a great advantage.
People often underestimate how important this is to customers. The stability and maturity of the hypervisor is so much more important than any other system software product. If you have an operating system and it blue screens, it'll drag down one application. It's not a great thing to have happen, but in the grand scheme of things it's not a complete disaster. If a hypervisor blue screens and goes down, it'll drag down seven, maybe eight, applications. That order of magnitude is a bigger problem.
Secondly, in terms of architecture ESXi is a next generation hypervisor with a dramatically smaller footprint than Hyper-V; we're talking about 50 to 60 times smaller. Why does that matter? The more lines of code you have in a software product, the bigger the surface exposure for security attacks. [Our] greater stability and maturity also relates to the fact that our hypervisor is a lot slimmer, has fewer lines of code and does only what a hypervisor is supposed to do. Microsoft on the other hand has an entire OS which is part of the architecture of the hypervisor. You don't want that.
We have a lot of functionality that ensures higher consolidation ratios as well. We talk about some of the tricks that we can do in terms of memory management, for example, that ensures that you can load a hypervisor with more VMs than Hyper-V can.
Another important fact is that now that hypervisors are the new platform for data centers, as it is the very first thing that installs on top of the hardware, the platform is only useful if it can interoperate with anything and everything else in the data center. Hyper-V can only be used with Windows 2008. Ironically, VMware supports more versions of Windows than Hyper-V does. And this is another reason why the immediate usefulness of Hyper-V is limited.
What about features?
Balkansky: So far I've only just touched on the hypervisor to hypervisor comparison as it is the only far comparison that can be made. We haven't even talked about the other things that VMware provides on top of the hypervisor that Microsoft isn't even thinking of, or has just given up on this time around. Last year they were talking about live migration as being part of the first edition; they ended up ditching that.
The second layer above the hypervisor, the virtual infrastructure capabilities, are the set of capabilities that aggregate multiple servers in a resource pool. These are the set of capabilities that act across an entire farm of servers. This is what really unlocks the true power of virtualization. Being able to partition a single server in a very static way is useful, but if you really want to leverage the power of virtualization you really need capabilities like live migration, dynamic load balancing and automatic restart of virtual machines (VMs).Only then you can really squeeze out more efficiency in terms of capital expenditures.
These are the kinds of capabilities that lead you down the path of automation that translates into saving on operating costs and simplifying management of your IT. More than 60% of our customers use VMotion in production. This is a must have requirement for virtualization. It's not some bell and whistle or a luxury. It's a must have for our customers.
How does CitrixXen come into the picture?
Balkansky: It doesn't [laughs]. I'm just being . . . Citrix reported their numbers for Q1 and reported $34 million [GAAP] from the XenSource product line. We reported $438 million in the same quarter. The thing about XenSource is that it has been on the market for a long time. You'd expect that the acquisition by Citrix would have given them a huge boost [by way of] access to a distribution network. But the uptake simply is not there.
I think virtualization is a classic example of a platform market. Any platform needs to interoperate with many things and is useful only if it does this. If you're a systems vendor, storage vendor or an ISV, you need to look at how you're going to certify your product and make it interoperate with these virtualization platforms that are coming out. You look at the market and you have VMware, the market leader with overwhelming market share for the time being; there is Microsoft, which is coming to market; then there are the others. And everyone has to place bets.
Doing certification and testing, while at the same time investing in those things, is very expensive for vendors. In all likelihood, I think that most of the ecosystem is going to place very few bets. I also think that so far they're placing their bets on VMware because of its 100,000 customers and the richness of the market for building products in and around VMware.
Obviously, Microsoft has tremendous market power but if I am a member of the ecosystem, these are probably the bets I'm going to place. For the rest of the contenders on the virtualization market, XenSource included, I think they're going to find it very difficult to compete. I don't think XenSource has really managed to offer something to differentiate itself so far.
What about their leverage on the desktop side?
Balkansky: Citrix obviously has a customer base on the desktop side and Presentation Server has a big footprint. I think that people try to equate the two: Because Citrix has a lot of desktop customers that therefore means that somehow they're going to become successful. I think that this really misses the point and overlooks the dynamics of the market at present.
People miss the fact that this whole interest and movement towards virtualizing the desktop is primarily driven by migrating the desktop functionality back to servers--back to the data center. This movement is very much being seen as an extension of the data center. Imagine the dynamics: At the end of the day you decide that desktops are going to be running on servers back in the data center. That's a data center decision that is, in many cases, made by the guy who runs the data center and runs the servers.
I doubt if the desktop group is going to have the power to tell the data center and server guy that the desktop VMs they're going to install are a different virtualization product. They've had VMware, but now they're going to install Xen Citrix because they want it that way.
The desktops are running in the data center, they're managed with the data center management tools and they're running on the servers. Everyone wants to standardize their infrastructure with the least amount of technologies as possible; the fewest number of vendors. If you're going to make a decision about virtualization platforms in the data center, you're going to make one decision. It may not be VMware in every case. But you're going to make one decision, not five decisions. For this I'll have this, for that I'll have that, for this . . . and so on. For that reason, I don't think that Citrix is going to be able leverage their position in the marketplace.