The Open Road

Virtualization may offer a significant advantage to Linux in the decade-old debate over Linux vs. Windows total cost of ownership (TCO). A new Gabriel Consulting Group survey (PDF) of mostly mixed-environment (that is, Windows and Linux) enterprises reveals significantly higher adoption of virtualization technology, with all the cost savings that go with it: less money spent on hardware and licensing fees.

It's an interesting conclusion, but leads to an even more interesting question: why don't Windows administrators take advantage of virtualization to the same extent as Linux administrators? The answer--licensing cost and complexity--is something that Microsoft has the ability, but not the interest, to change.

According to the survey, enterprises that predominantly use Linux virtualize roughly 30 percent more than those that prefer Windows, and heavier virtualization users do so much more aggressively on Linux systems than on Windows:

Linux vs. Windows: Virtualization Trends

(Credit: Gabriel Consulting Group)

The survey's author reports that "Linux users have clearly both adopted virtualization at a greater rate and embraced it to a greater extent than customers who have standardized on Microsoft operating systems," but why?

Perhaps the primary reason is that Microsoft didn't really start to promote virtualization until long after the Linux crowd. This isn't surprising: Microsoft has much to lose from virtualization. The fewer Windows server licenses an enterprise has to buy, the worse it is for Microsoft.

Microsoft has now jumped into the virtualization market with both feet, giving its Hyper-V product away for free...but not really. Indeed, it is the pricing strategy Microsoft has for its servers that may go furthest in explaining its lack of appeal to Windows users, as noted in Gabriel Consulting Group's report:

There are also licensing differences that bear directly on comparative costs. With Microsoft, users who don't have volume agreements or who haven't purchased the more expensive Enterprise or Datacenter editions will have to purchase licenses for every system and each of the virtual machines running on those systems. Linux, on the other hand, can be essentially free, meaning that companies can deploy it on multiple systems or in virtual machines at no cost.

While the survey also lists the benefits of source code access to Linux administrators, I suspect that this is of minimal value to the big majority of Linux adopters. Very few will care to "get intimate with the code," to use the report's language, preferring instead to stick to the more tangible (and easily accessed) cost savings from Linux virtualization.

There are other benefits to those who primarily adopt, or standardize on, Linux, as the report suggests:

• 77 percent of survey respondents reported greater hardware utilization rates through Linux virtualization, versus 56 percent of Windows users.
• Those who standardize on Linux find Linux virtualization much more manageable (62 percent) than Windows administrators who standardize on Windows virtualization (48 percent). More telling, four times as many Windows standardizers (23 percent) find Windows virtualization hard to manage than the Linux standardizers, only 6 percent of whom find Linux virtualization hard to manage.
• Linux translates into higher server utilization and, hence, less power consumption and more physical space: 59 percent of Linux administrators disagreed with the "We are rapidly running out of data center electrical capacity" statement, compared to 38 percent of Windows administrators. When presented with the statement "We are rapidly running out of data center floor space", 60 percent of Linux administrators disagreed versus 45 percent of Windows administrators.

While enterprises could realize even bigger cost savings by simply using free Linux versus paid Windows, most enterprises will buy commercial support for Linux through Red Hat, Novell, or Canonical. Even factoring in this cost, however, Linux seems to lend itself more readily to virtualization and, hence, to cost savings that result therefrom.

Microsoft has it in its power to turn the tide relative to Linux's superior virtualization TCO, and it probably has little to do with the cost of Windows Server, and certainly not with the cost of its Hyper-V virtualization technology, which is now $0.00.

Rather, it's likely a matter of simplifying its famously Byzantine pricing, and making Windows Server licensing friendlier to virtualization. For example, Microsoft doesn't allow migration of its products to a new physical server more than once every 90 days. This may ensure customers buy licenses with fewer restrictions, but it also appears to mean they simply buy fewer Microsoft licenses, period.

Given that commercial Linux isn't free, Microsoft doesn't need to make Windows free to make its Hyper-V virtualization more competitive with Linux virtualization. Simplification, it seems, would go quite far toward the goal of making Windows virtualization more palatable.

Goldman Sachs IT Spending Survey

(Credit: Goldman Sachs)

IT spending may be tight, but chief information officers plan to increase their budget allocation to a select group of virtualization vendors, including VMware, Citrix, and Red Hat, according to a Goldman Sachs CIO survey released Monday.

It's not surprising that virtualization is top of mind and wallet for CIOs, but things look particularly rosy for Red Hat, given its position as the market leader in open source and a strong challenger in virtualization.

While the percentage of CIOs expecting to increase IT spending has grown since Goldman Sachs' last survey in June 2009, a full 69 percent expect to maintain or decrease their IT spending.

Against this backdrop, Goldman Sachs sees Red Hat boosting its share of IT spending as the open-source leader claims the lion's share of a Unix-to-Linux server shift that "remain(s) in the early innings." Equally important, Red Hat is seen as a critical integration and distribution point for other vendors:

Red Hat is positioned well for the emerging cloud-computing ecosystem, given its open-source background and current positioning in data centers, including enterprises as well as cloud providers such as Amazon. In addition, Red Hat's strategic importance to others is also increased by its platform capabilities that provide a beachhead for many other software products into the corporate data center. That being said, cloud computing remains a nascent opportunity with little revenue contribution to date and an increasing competitive landscape.

To date, Red Hat has mostly resisted the temptation to expand its product portfolio beyond the operating system, and directly adjacent opportunities like virtualization and cloud computing. However, as the company further strengthens its balance sheet and grows in confidence, we should finally see Red Hat use its dominant brand to give CIOs more reasons to pay Red Hat money.

(Credit: Goldman Sachs)

Intriguingly, Red Hat may be pushed to this step by the increasingly ambitious VMware, which has far more cash and a strong interest in being the foundation for enterprise's cloud-computing technology. According to the Goldman Sachs report:

VMware is a leader in three important growth themes in IT: server virtualization, desktop virtualization, and cloud infrastructure. We also believe that as virtualization penetration increases, the company has an opportunity to take significant share of the large systems management software market. Microsoft's increasing focus on the space is a risk; however, our latest checks give us greater confidence in VMware's customer loyalty and the company's significant technology lead. We also see room for significant ongoing margin expansion as the company matures.

With the recent acquisition of open-source vendor SpringSource, VMware can deliver on the powerful "Build-Deploy-Manage" mantra that SpringSource championed to its 2 million developers.

Both companies should thrive as IT budgets remain lean. But which will ultimately benefit most is a question of execution and ambition.

Follow me on Twitter @mjasay.

Old dogs may struggle with new tricks, but they seem to be able to figure out new licenses.

In a shocking move, Microsoft announced Monday the release of Hyper-V Linux Integration Components (LinuxIC).

The news reflects Microsoft's continued interest in lobotomizing its virtualization competition through low prices, but also the recognition that it must open up if it wants to fend off insurgent virtualization strategies from Red Hat, Novell, and others in the open-source camp.

But the truly startling news is that LinuxIC is being released under the GNU General Public License (version 2). Microsoft once called GPL anti-American. Now it calls it friend.

The gods must be crazy.

Or maybe Microsoft is simply recognizing (finally!) that GPL can be a capitalist's close ally. That and the fact that many components within the Linux kernel are GPLv2-licensed make the move completely natural...at least, once you forget that this is Microsoft embracing GPL, rather than some other company like Red Hat.

LinuxIC is a collection of kernel drivers that enable Linux to recognize that it is running on Microsoft's Hyper-V and optimize accordingly, resulting in an "enlightened version of Linux," according to market researcher IDC. The device drivers have yet to be accepted into the Linux kernel, but the GPL license and general utility makes their inclusion probable.

The move opens up Hyper-V to much more than Windows, which has arguably been its weakest point. As IDC notes, this embrace of Linux is a "key element if Microsoft is going to successfully go head to head with VMware in large accounts--many of which already are dedicated VMware customers."

Importantly, Microsoft is now opening up even beyond its long-time Linux partner, Novell, to embrace an array of other Linux partners, including Red Hat. While Novell was the first Linux vendor to certify for Hyper-V, Microsoft's lack of real support beyond Novell's Suse Linux Enterprise Server was a weakness, as some have complained.

But this is arguably a new Microsoft. Redmond recently announced that Office 2010 will support Internet Explorer and Mozilla's Firefox. The company is learning that its customers run heterogeneous software environments, and it's (slowly) responding. Microsoft's Sam Ramji, senior director of Platform Strategy, notes: "We are seeing Microsoft communities and open source communities grow together, which is ultimately of benefit to our customers."

Microsoft, in short, can't ignore open source, including Linux, without ignoring its own customers.

But surely this move is more Machiavelli than Santa Claus? Maybe, maybe not. I asked Novell's Greg Kroah-Hartman, a prominent Linux kernel developer who was deeply involved in influencing Microsoft to release LinuxIC, what Microsoft's move means for Linux. His response reflects an enthusiasm that is as surprising as it is refreshing:

We want Linux to work well for everybody. This move is not bad in any way for Linux, Xen (Novell's preferred virtualization technology), or KVM (Red Hat's preferred virtualization technology). This is not a competition, per se.

With LinuxIC, Microsoft is doing two things. First, it's saying that contributing open-source software under GPL is acceptable. And second, it's supporting the idea, which I and others in the Linux kernel community have long advanced, that all Linux kernel drivers should be open source.

LinuxIC is the latest example of how Microsoft is changing, and it's a big proof point. When Microsoft embraces Linux, that's news. When it does so by embracing GPL, it's perhaps time to start the countdown to Armageddon.

Follow me on Twitter @mjasay.

Could VMware be the next Novell? That's the question Gartner managing vice president and chief of research for Infrastructure David Cappuccio asks in a provocative post, one that bears further discussion. While VMware is at the top of its game, there are several historical analogs between VMware and Novell.

I'll let you read Cappuccio's excellent post for his full argument, but the crux of it is that in the face of dominant but pricey technology, many buyers will turn to "good enough" to fill their needs. For Novell, that competition to its 90 percent market share came from Windows, which displaced Novell's "great technology that was more complex (or complete) than most customers needed."

Today, VMware faces a host of rising threats. Cappuccio picks out Microsoft's Hyper-V as chief among them:

[L]urking in the background is this little thing called Hyper-V; not as robust, or as tested as VMware, with almost no install base, and certainly not ready for prime time in most people's minds. However, it will be an integral part of Windows 7, Windows Server 2008 and Windows Server 7 in 2010. Why should you (or VMware) care? Because like "free networking", or "free SharePoint", hyper-V will get used, slowly at first, but as more and more systems get installed the base will increase and within just a few short years companies will discover (surprise, surprise!) that they have business applications running on both VMware and Hyper-V.

Free-and-good-enough is a great strategy, and one that Microsoft has long used to exceptional effect.

Of course, Microsoft isn't the only one playing this game. Xen is included for free in Linux, though Red Hat is pushing to move users to KVM (and succeeding to an increasing extent). Virtualization customers are spoiled for choice.

All of which leaves VMware exposed. This isn't to suggest that VMware should resign itself to obliteration. Indeed, VMware has gone on the offensive, releasing a host of software as open source to combat open-source alternatives, most intriguingly its open-source virtualization client, as OStatic's Sam Dean notes.

Novell didn't respond to its Netware demise until it was too late. VMware seems to be learning from history.

The real question is whether it will be able to go as low as Microsoft and the Linux vendors on price while still maintaining "good enough" profits. I suspect it will fail in this because for VMware, virtualization is core and it must price accordingly. For Microsoft, Red Hat, and Novell, virtualization is a critical complement, but not the core of their businesses. Complements are cheap, core is not.

Follow me on Twitter @mjasay.

While evangelists of Linux distributions built for personal computers (i.e., "Linux desktops") point to Netbooks as an indication of renewed life in their chances to compete for consumers, new data suggests that this may be a fool's hope.

Instead, such advocates would do well to follow the leads of Canonical and Red Hat, as they respectively extend the desktop with cloud services and deliver desktop functionality from the cloud.

Although it's true that roughly 30 percent of Dell Inspiron 9s Netbooks run Ubuntu Linux, it's equally true that about 90 percent of Netbooks run Windows, as Computerworld recently pointed out, while Linux had started with 30 percent of the Netbook market.

And while Dell is now saying that the return rate on Linux Netbooks is no longer four times that of Windows, as it was reported in October 2008, it's unclear how or why Linux will be able to take a greater share of the Netbook market, given that Microsoft has reduced its pricing to compete with Linux.

Sure, this means lower profit margins for Microsoft, but that's a hollow victory for Linux, isn't it?

Linux has a much better chance of succeeding on personal computers, if it starts from a position of strength, not weakness. Two areas of strength for Linux are mobile and servers. The mobile Linux market, however, remains somewhat fragmented, making it difficult to mount a near-term desktop challenge.

The Linux server, however, is ripe to creep down into the desktop, and that's precisely what Canonical (Ubuntu) and Red Hat are doing.

The two companies are going in opposite directions, but they end in similar positions. For Canonical, which, with Ubuntu, arguably has the strongest claim to innovation and leadership on "the Linux desktop," the trick is to move more personal-computing services into the cloud.

Canonical's Mark Shuttleworth told me last year that his Ubuntu desktop strategy would increasingly include cloud services. Recently, Canonical started to deliver on this vision by Amazon EC2 (Elastic Compute Cloud) hooks to its server edition, which has the added bonus of giving Canonical a compelling revenue model.

Red Hat, for its part, is starting with the server, where it's the undisputed market share and value leader. Recently, Red Hat told Computerworld that it plans to grow its so-called desktop footprint with a "desktop" that isn't: it's a virtual machine running remotely on a server and sitting side by side with Windows.

There may be flaws in this strategy, as some have pointed out, but as Red Hat CTO Brian Stevens told me on Wednesday, Red Hat's approach makes a lot of sense for CIOs who want to increase manageability of desktops while simultaneously reducing costs:

The target we are designing for is not the legacy model of thick Windows clients or terminal services. Open source is about driving innovation and new paradigms of use, not just to make a cheaper alternative to proprietary (software).

So the desktop target model we are designing for has several elements:

• Enable open-source solutions to help lower the OpEx costs of Windows environments by allowing Windows desktops to be virtualized within the data center;
• Open up the interoperability and technical advancement of the desktop remote-protocol space by open sourcing the Spice protocol for building VDI infrastructure on. This will have impact not just on virtual desktops, but also remote display of physical desktops;
• Through our ovirt.org effort, build an open-source implementation and reference architecture for building clouds, on which servers and desktops can be instantiated on demand;
• Bring virtualization to the client through our KVM technology, putting a hypervisor directly on the desktop. Enable the client to not just plug into VDI environments, but to be able to run desktops within the cloud or locally, seamlessly, and with full mobility;
• Through this new model, which enables multitenancy of desktop environments, enable a virtualization-optimized Red Hat Enterprise Desktop to be run from the same pane of glass as Windows.

In sum, Red Hat is not merely looking to join the Linux game for personal computers; it's also seeking to completely change the desktop market--Linux, Windows, or otherwise. This is an ambitious goal--one that Red Hat's Linux server strength puts it in a good position to deliver.

Linux desktop advocates should take a cue from two leading Linux companies, Red Hat and Canonical. The point isn't to replicate the Windows desktop. The point is to completely change the way desktops are delivered and their services consumed. Anyone still worried about Linux on Netbooks is fighting the wrong battle.

Follow me on Twitter at mjasay.

For years, Microsoft has insisted that open-source vendors acknowledge that its patent portfolio is a precursor to interoperability discussions. On Monday, Microsoft shed that charade and announced an interoperability alliance with Red Hat for virtualization.

The deal includes several key components, all related to virtualization:

• Red Hat will validate Windows Server guests to be supported on Red Hat Enterprise virtualization technologies.
• Microsoft will validate Red Hat Enterprise Linux server guests to be supported on Windows Server Hyper-V and Microsoft Hyper-V Server.
• Once each company completes testing, customers with valid support agreements will receive coordinated technical support for running Windows Server operating systems virtualized on Red Hat Enterprise virtualization, and for running Red Hat Enterprise Linux virtualized on Windows Server Hyper-V and Microsoft Hyper-V Server.

Pretty straightforward, as interoperability should be, and driven by customer demand for Microsoft technologies running alongside Red Hat's, according to Mike Neil, general manager of Virtualization Strategy at Microsoft. The top Linux vendor partnered with Microsoft: this is a major win for customers.

Crucially, Red Hat's interoperability deal with Microsoft does not include any patent covenants, the ingredient that torpedoed Novell with the open-source community:

The agreements establish coordinated technical support for Microsoft and Red Hat's mutual customers using server virtualization, and the activities included in these agreements do not require the sharing of IP. Therefore, the agreements do not include any patent or open source licensing rights, and additionally contain no financial clauses, other than industry-standard certification/validation testing fees.

Red Hat has long argued that patent discussions only cloud true interoperability, which is best managed through open source and open standards.

While Red Hat has flirted with such interoperability before by joining with Microsoft in the somewhat toothless Vendor Interop Alliance, this is its first direct interoperability initiative with Microsoft.

What most people don't know is that Red Hat had been discussing interoperability initiatives with Microsoft for a year before Novell and Microsoft tied the knot, but Microsoft ultimately derailed the talks by trying to introduce a covenant not to sue over patents, similar to what it ended up negotiating with Novell. Red Hat rejected this unnecessary inclusion, left the bargaining table, and Microsoft connected with Novell to use interoperability as an excuse to attack open source.

Monday, Red Hat and Microsoft have together demonstrated that interoperability can exist independent of back-room dealings over patents. Microsoft has increasingly been forced to open its stance on patents by the European Commission, anyway, proving Red Hat's resolute stance against patents was the right one. But this announcement suggests that Microsoft is maturing in its views on how to interact with open-source vendors.

It also suggests that Red Hat is maturing in its realization that it must interoperate with the old world of proprietary software even as it attempts to forge a new one of open-source software. Red Hat has long depended upon proprietary software: Red Hat Enterprise Linux's success has derived from its support for Oracle and other proprietary vendors.

Both Red Hat and Microsoft on Monday lowered their guns long enough for customers to win. They did so without encumbering interoperability with patents, which will be critical to ensuring that Microsoft can lower its guard further to welcoming open-source solutions to the Windows fold as a full partner.

Follow me on Twitter at mjasay.

There hasn't been much Web chatter around Open Kernel Labs, but late last week, the Chicago-based Open Kernel Labs, a spinout from Australia's NICTA, announced a $7.6 million investment from Chrysalis Ventures, Neo Technology Ventures, and Citrix Systems.

This follows a $2.5 million grant Open Kernel Labs recently received from NICTA.

Not much noise is made about Open Kernel Labs because it operates in the embedded-virtualization market, providing microkernel technology to manufacturers of electronics such as mobile handsets.

Importantly, while based on the open-source General Public License 3, the company is able to segregate differently licensed components and run on Linux, Windows, or other embedded operating systems (including real-time operating systems), which is a critical requirement in the embedded world.

True to its open-source roots, developers can dig through information on the source code, another key advantage to an open-source microkernel technology. The embedded market invests a lot of time and resources in customizing code: it's the one market in which modifiability of source code is a must-have feature.

I haven't seen much funding in the embedded-operating-system market since my own company in that market, Lineo, sold to Metrowerks in late 2002. It's nice to see the eruption of the mobile market turning interest toward open-source embedded software again.

Installing an open-source enterprise application has never been easier. No hardware? No sophisticated IT department? No problem. At least, not if you use one of 38 JumpBox-enabled open-source applications, as it announced recently.

A rising number of companies offer virtualized instances of popular open-source applications, but JumpBox takes it a step further, deploying to the Amazon Elastic Compute Cloud (EC2) service, almost completely obviating hardware and setup quandaries.

JumpBox offers small to midsize organizations a library of open-source applications packaged as pre-built, pre-configured virtual appliances through JumpBox Open, its annual subscription service. Public Amazon Machine Images (AMI) for 12 JumpBox applications, including Ruby on Rails, (Alfresco, Movable Type, Magento), Drupal, SugarCRM and more have been made available for free. AMIs for the full suite of 38 applications are available to plus and premium subscribers to JumpBox Open.

Pricing of JumpBox Open starts at $299 per year (for one persistently running JumpBox instance of each application), rising to $999 per year to run up to 15 simultaneous production instances of any JumpBox-enabled application. In other words, it's dirt-cheap.

Powerful software, low price, and no fuss. What's not to like? If you're an SMB customer, probably not much.

But if you're an open-source application vendor, I suppose it's still an open question how JumpBox will work with you to share revenue. In my conversations with the JumpBox founders, this potential conflict has come up, and I know the JumpBox team is working on it. How well it gets resolved may well determine how much emphasis open-source vendors will put on the JumpBox sales channel which, in turn, could decide the fate of JumpBox.

With or without the vendors, however, this is a great service and suggests a bright future for enterprise software.

Disclosure: I work for Alfresco and advise several of the companies whose open-source applications JumpBox distributes.

VMware may still hold 78 percent of the virtualization market when measured by revenue, as reported by ComputerWorld. But Microsoft has learned a thing or two from open source and now claims a whopping 23 percent of the market based on shipments of new licenses, with VMware down to 44 percent of new licenses.

With just a few months in the market with its Hyper-V product, Microsoft's market share progress signals discontent with VMware's pricing and suggests that Microsoft may not be the only one capable of feeding at the virtualization trough.

Red Hat and Novell: time to kick your virtualization efforts into overdrive. VMware's customers await you.

Microsoft's progress has been made under the banner of low cost and choice. As noted by Microsoft, "Customers now have choice in market. VMware is no longer the dominant server virtualization vendor." It's a line stolen from the open-source script, and one that Red Hat and Novell would do well to use.

Red Hat and Novell aren't even in the top-three vendors for virtualization shipments, as IDC data indicates. Still, Red Hat has been charging hard into virtualization with its Qumranet acquisition, and both Red Hat and Novell have something that even Microsoft can't match: a free entry price for using their virtualization offerings.

The one sure thing in all this virtualization fracas is that it's a bad time to be offering virtualization as a standalone. The operating system vendors are going to win this battle because they can bury the cost of virtualization in larger offerings.

VMware came into the stock market with a bang last year, but it may go out with a whimper.