History

When Microsoft officially officially launched its new Server Virtualization Validation Program (SVVP) just two weeks ago, an unexpected vendor applied for it: Cisco.

The SVVP program is developed to extend the support that Microsoft offers on Windows Server and most back-end servers to those 3rd party hypervisors that pass validation.

As far as we can understand, there’s no reason to apply for this program without a hypervisor.
So, or Cisco is about to announce a hypervisor, or the SVVP has additional purposes that Microsoft didn’t detail.

Today another unexpected vendor appears on the SVVP applicants list: Unisys.

It’s not a secret that Unisys is heavily involved in virtualization, omnipresent in virtualization adoption projects across the globe, but as far as we know the company doesn’t offer (yet) a hypervisor.

As Unisys contributes to the Xen open source project, this may be the first step to launch a Xen-based solution.

Just two weeks ago Microsoft officially launched its new Server Virtualization Validation Program (SVVP), developed to extend the support to those 3rd party hypervisors that pass validation.

At the announcement time a number of key virtualization vendors already applied for the Program, including Citrix, Novell, Sun, Virtual Iron and Cisco (why Cisco?).

At the last minute VMware confirmed its commitment and now, ironically enough, its flagship hypervisor, ESX 3.5 Update 2, is the first to pass the validation.

This means that from now on, Microsoft had to deliver cooperative technical support to those VMware customers running Windows Server 2008 and earlier, as well as any Microsoft back-end server defined in the program (and this includes mission critical products like SQL Server and Exchange).

With a surprising announcement Novell unveiled its plans to enter the application virtualization market yesterday.

The company signed an OEM agreement with a small firm called XenoCode, which recently reshaped its technology to deliver application virtualization.

Novell already rebranded the XenoCode Application Studio as ZENworks Application Virtualization (ZAV), offering it at $39 per concurrent user.

With this move Novell confirms its strong commitment on virtualization.

Novell already has its own hypervisor, the Xen implementation embedded in SUSE Enterprise Linux 10, but it’s also planning a second virtualization platform featuring just the hypervisor.

On the management side its ZENworks Orchestrator acts as hypervisor console, VM lifecycle management and virtual infrastructure orchestrator.

Last but not least, at the beginning of the year Novell acquired PlateSpin granting itself P2V migration and capacity planning capabilities that should be integrated in the ZENWorks family by the end of this year.

Sun releases today the second generation of its desktop virtualization platform, VirtualBox, acquired by innotek in February.

This new version introduces a few but important capabilities like:

• support for 64-bit guest OSes (Microsoft Windows Vista and Red Hat Enterprise Linux 5)
• support for AMD RVI
• support for Microsoft VHD virtual disk format
• new interface for Apple Mac OS edition (native 10.5 Leopard GUI)
• Python API (for Solaris and Linux hosts only)

Download it free of charge here.

The virtualization.info Virtualization Industry Roadmap has been updated accordingly.

Last week Citrix announced that the its next generation desktop and application virtualization platform, XenApp 5.0, will be launched on September 10.

The company made the product available slightly ahead of time so you can download a trial later today here.

Citrix also published a valuable Technical Guide for Upgrading or Migrating to XenApp 5 that every customer should read.

The virtualization.info Virtualization Industry Roadmap has been updated accordingly.

Xen 3.3 has been just released and Xen.org already publishes the roadmap for the next version of the open source hypervisor.

The document reveals a major focus on desktop deployment, with multiple major enhancements to make Xen a great client hypervisor:

• Client device virtualization (e.g. battery status etc)
• GPU virtualization with Gallium
• USB 2.0 support, PV USB support
• Trusted HID
• Simple VGA/text-mode management console

Probably, the most important entry in the list above is the display card (GPU) virtualization, one of the most complex task to achieve in hardware virtualization.

As reported by Wikipedia:

Gallium3D is a software library for 3D graphics acceleration being developed by Tungsten Graphics, an engineering company with expertise in Linux and open-source graphics technologies. Gallium 3D operates between the graphics API and the operating system with the primary goal of making driver development easier, bundling otherwise duplicated code of several different drivers at a single point.

…

Gallium3D provides a unified API exposing standard hardware functions such as shader units found on modern hardware. Thus, 3D APIs such as OpenGL 1.x/2.x, OpenGL 3.x, OpenVG, GPGPU infrastructure or even Direct3D (as found in the Wine compatibility layer) will need only a single back-end, called state tracker, targeting Gallium3D API…

To allow a full GPU virtualization, display cards vendors (Intel, AMD/ATI and nVidia mainly) will have to support the technology.
The current status reported is not exactly encouraging:

The first implemented and already partially working drivers are Cell and Intel GPU drivers. Work is done on ATI Radeon cards providing a skeleton driver, and the Nouveau team is moving development to Gallium3D, including a solution for older fixed function nVidia cards which lack programmable shaders.

Several companies are currently working to make Xen a client hypervisor: the BIOS leader Phoenix Technologies (with its HyperCore), the just-launched startup Neocleus, and several major OEMs like Dell, HP and Lenovo.

If Xen 3.4 will really develop the features above these companies may get major benefits from it and push for an early implementation in their commercial products.

The upcoming new version anyway is not just about desktops. It includes an impressive range of additional features, including:

• full support for Intel VT-c (VMDq and SR-IOV)
• full support for OVF
• Virtual machine synchronization for fault tolerance
• SATA command virtualization

In June 2007 Google acquired a stealth-mode startup focused on application virtualization called GreenBorder.
As common practice, the search giant never detailed how it planned to use the technology.

Yesterday the company unveiled the first public beta of its own browser, Chrome, featuring the capability to isolate the instances running in each tab.

InformationWeek is reporting that this security feature depends on the GreenBorder technology.

While there’s no official confirmation, it makes sense and leaves open a wide range of possibilities: if Google really deploys an application virtualization platform along with its browser, it could stream inside the virtualized layers any kind of application, on any kind of operating system.

By the way: virtualization.info is fully compliant with the Chrome rendering engine, so feel free to browse us with it.

In the past few weeks we introduced a number of top-notch speakers that will give their keynotes at the Virtualization Congress 2008, the independent conference that virtualization.info will hold in London, October 14-16.  

All of them are well-known protagonists in the virtualization industry and really don’t need any introduction. Except one: Mark Russinovich, Technical Fellow at Microsoft.

For those ones that don’t know him, Mark is a legendary figure in the IT world because of his previous companies: Sysinternals and Winternals (both links now redirect to Microsoft TechNet).

While working there, Mark did so much reverse engineering of the Windows kernel that he ended up knowing it as much as the Microsoft architects. And this is the reason why Microsoft acquired his assets two years ago and hired him as a Technical Fellow.

The tools that he developed during the last ten years became fundamentals for generations of IT professionals in understanding the Windows behavior, and are still unrivaled at today.

Now what does Mark have to do with Microsoft virtualization?

As Technical Fellow, Mark is deeply involved in the development of the Windows kernel (Windows 7 and beyond), taking care that its architecture is fully virtualization-aware.
If there’s one that can have a discussion about virtual resources hot-plug, memory overcommit, performance comparison against VMware ESX and other virtualization hardcore topics, Mark is that one (this recent interview is a clear example).

Best of all, if you check his keynote abstract you’ll discover that Mark will preview some unannounced products at the Virtualization Congress, giving a glimpse of the Microsoft effort in the space.

Here’s Mark introducing his own keynote:

The Virtualization Congress 2008 is next month, October 14-16 at the London ExCeL Conference Centre.

Register now!

It’s well known that multiple companies are working to deliver connection broker capabilities to the new Microsoft Hyper-V 1.0. The list includes at least Citrix, Quest/Provision Networks, Ericom and VDIworks.

So far anyway was unclear how these company interact with Hyper-V and how complex it could be for newcomers.
Now Microsoft clarifies that it exposes a new set of APIs for the Terminal Services Session Broker to achieve the task:

…while the built-in functionality of the TS Session Broker in WS08 only supports routing connections to terminal server sessions, we’ve created a set of APIs that ISVs can use to create connection brokers for other kinds of devices. Basically, these APIs allow you to lobotomize the TS Session Broker and replace its brain—its brokering mechanism—with a new plug-in. This plug-in can contain a new set of rules that support redirection to other types of destinations. It can also provide different means of deciding the best target for new connections, such as load balancing rules based on server resources or login time…

The existence of these APIs could simplify the entrance of new competitors in the VDI market, which is becoming hotter and more crowded every quarter.

As multiple sources are reporting, Alex Vasilevski, the co-founder of Virtual Iron, is about to launch a new startup temporary called Old Road Computing.

Vasilevsky left Virtual Iron in December 2007 to immediately found this new firm with Dan McCall.
McCall comes from Reflex Security, one of the first security firm approaching the virtualization market in early 2006, and Verisign.

The new company is currently funded by Highland Capital Partners and Flybridge Capital Partners (formerly IDG Ventures) for an undisclosed amount.

Mass High Tech reports that Old Road Computing will be renamed as Virtual Computer Inc. as soon as the team is ready to leave the stealth mode.

Old Road Computing (until there will be a confirmation of the new name) has been included in the virtualization.info Virtualization Industry Radar.

Update: Old Road Computing officially confirmed its new name: Virtual Computer.
The virtualization.info Virtualization Industry Radar has been updated accordingly.