virtualization.info

Last week Microsoft announced some remarkable changes in its VDI strategy, including new licensing options and the upcoming integration of Calista technology in Remote Desktop Services (RDS) under the name of RemoteFX.

The announcement, and the fact that Microsoft highlighted a stronger-than-ever partnership with Citrix in this effort, will impact the VDI ecosystem and, by some degrees, the customers adoption of desktop virtualization technologies.
The first reactions already came in: VMware, Quest and WYSE already published some comments on RemoteFX.

VMware, threatened by the “Rescue for VMware VDI” promotion, has a lot to say of course and lists the downsides of the offering:

• The trade-down includes two different product suites from two different companies with overlapping components.  Both solutions come with a hypervisor platform and connection broker.  This creates confusion leaving the customer asking which components to use when?
• A set of non-integrated products marketed as a solution with no reference architecture for deployment
• The included licenses are actually only free for one year (on the Citrix side) and then you’ll need to pay for that license + SA in subsequent years.
• Hidden costs of more infrastructure, more time and increased complexity to deploy and manage

A further elaboration on the complexity aspect would be greatly appreciated, but the other points are clear and the first one is very relevant: the overlap of Microsoft and Citrix virtualization offering always created confusion and the two partners never did a good job in providing guidance about this aspect.

VMware also has a long list of comments on RemoteFX:

• With the release of RemoteFX, Microsoft has acknowledged that the best way to deliver a rich user experience is via a protocol, like PCoIP, designed for effective host side rendering.

• Citrix joins in the acknowledgment by agreeing to support RemoteFX alongside their own “HDX” and in turn points to some of the deficiencies of ICA.
• On the management side, only VMware View offers a single centralized management solution. Other vendor solutions, such as Citrix XenDesktop and Microsoft RDS, use a myriad of non-integrated tools, consoles, and wizards to perform even the most basic tasks.
• RemoteFX will be a LAN-only solution (much as RDP is today). This is more limited than the PCoIP protocol which was designed to provide a great user experience on both local and wide area networks.
• While Microsoft spoke of vendors providing both RFX enabled thin clients and server offload cards, so far all that has been provided is a reference design – no vendors have publicly committed to delivering these components. By contrast, there are already a number of vendors delivering PCoIP zero clients and PCoIP monitors including Dell, IBM, Samsung, and Wyse. Furthermore, Teradici has had hardware-enabled solutions on the host side for more than 2 years.
• RemoteFX does not treat the desktop as a composition of multiple types of content. Instead it treats each element of the desktop exactly the same. An advantage of the PCoIP protocol is that it has different CODECs to handle different image content such as text, graphics, and video. Since RemoteFX cannot recognize and transmit the desktop by content type, end-users need to settle for a lowest common denominator experience versus the high-definition environment the PCoIP protocol can deliver.
• |One of the key features of VMware View with PCoIP is the ability for the display protocol to dynamically adjust to changing network conditions. PCoIP can “build to lossless” , which means that the user will receive the best experience possible as network conditions change. Both ICA/HDX and RDP/RemoteFX “dial in” the user experience when a user first connects and have limited ability to dynamically adapt to changing network conditions or user interactions.
• The idea of virtualizing physical GPUs is quite compelling, but unfortunately it’s a “chicken and the egg” type of problem. Most of today’s enterprise class servers are not designed to accept high-end graphics cards. Additionally, the cost of the GPUs can be quite high (upwards of $4,000 each) while only supporting a small number of users per card.

Quest, which could be seriously damaged by the advent of RemoteFX and the renewed partnership with Citrix on VDI, even releases an official press announcement to clarify its commitment and explain where vWorkspace will add value to the new platform:

Quest Software, Inc., provider of vWorkspace, a single virtualization solution for application delivery and desktop deployments, today announced plans to support Microsoft RemoteFX, a Windows Server 2008 R2 Service Pack 1 feature currently under development, in future versions of Quest products.

…

Added Rolls, “With our vWorkspace product, we will continue to enhance and extend the Remote Desktop Services platform to deliver user experience optimizations, especially for remote workers with WAN or Internet connections. We believe these optimizations will be especially important for customers using RemoteFX, and currently plan to extend our support to include RemoteFX within 30 days of its availability.” …

Last but not least, Wyse Technology, one of the largest thin client vendor in the VDI/SBC space, mentioned by VMware in the comments above, releases a press announcement too, to clarify the support to RemoteFX in every future product:

Wyse Technology, the global leader in thin computing and client virtualization, today announced plans to support Microsoft RemoteFX technology, enabling rich media experience across all Wyse client platforms.

Microsoft RemoteFX, a feature being developed for Windows Server 2008 R2 SP1, will integrate with any and all Wyse devices and both companies expect today's announcements to accelerate desktop virtualization deployments…

Labels: Microsoft, Quest, VDI, VMware, Wyse Technology

Yesterday Microsoft announced a number of new technologies and initiatives around desktop/server virtualization and VDI.

The company also announced a new paper titled Virtual Hard Disk Performance.

The 35-pages document describes a benchmark executed by Microsoft to compare I/O performance of files inside its Virtual Hard Disk (VHD) format (both fixed size and dynamically created) against files inside raw disks and files inside the NTFS file system.

Tests were executed on systems running a number of different workloads, including SQL and Exchange.
Microsoft explains that compared to previous implementations, VHD support is native inside Windows Server 2008 R2 and thus is not depending on the presence of Hyper-V:

…the impact of the Windows hypervisor is quite small based on past experimental results. This is mainly due to the fact that performance critical workloads are re-directed to synthetic VMBus channels instead of using the longer emulation path. To get the most accurate CPU utilization and to focus on native performance, the Windows hypervisor is turned on only during VHD performance measurement in Windows Server 2008 which is required to mount VHDs on a Windows Server 2008 machine while it remains off for all other performance testing scenarios…

Microsoft decided to use a server (the vendors is undisclosed) with two quad-core Intel Nehalem-EP processors, 6GB RAM with NUMA enabled, serving 64bit Windows Server Enterprise 2008 and Windows Server Enterprise 2008 R2, attached to a Dell PowerVault MD1000 DAS.

The paper closes with an interesting table listing the pros and cons of different storage containers for Hyper-V. 

Labels: Benchmarks, Microsoft, Papers

One hour before starting a joint webcast with Citrix about its new virtualization strategy for desktops, Microsoft briefly announces a number of new initiatives, upcoming technologies and licensing changes.

About hosted desktop virtualization:

• The Windows XP SP3 virtual machine that can be run on Windows 7 Virtual PC, called XP Mode, will no longer require hardware virtualization to be executed.
  This is probably the best way Microsoft found to solve the problems that Sony created to many customers with its shortsighted strategy.
  
  The new version is available now as a hotfix.

About bare-metal server virtualization:

• Windows Server 2008 R2 Service Pack 1 will introduce a memory overcommit technique for Hyper-V R2 called Dynamic Memory.
  The news leaked at the beginning of February.

About VDI:

• The remote desktop acceleration technology acquired by Calista in January 2008, now renamed as RemoteFX, will arrive with Windows Server R2 Service Pack 1 and will be integrated in Remote Desktop Services (RDS).
  RemoteFX can be considered an accelerator for RDP over the LAN for Windows 7 SP1 clients only.
• Beginning July 1, 2010, Windows Client Software Assurance (SA) will include the VECD license for free.
  Customers that don’t wont to subscribe the SA will be able to buy a new Virtual Desktop Access (VDA) license: $100 /device/year instead of $110 of the VECD.
• Beginning July 1, 2010, Windows Client Software Assurance (SA) and new Virtual Desktop Access (VDA) license customers will have the right to access their virtual desktop and Office applications inside it on secondary, non-corporate network devices, such as home PCs and kiosks.
• Microsoft and Citrix signed a new technology agreement to integrate and extend Microsoft RemoteFX with Citrix HDX.
• Microsoft and Citrix launched a joint trade-in program dubbed “Rescue for VMware VDI”, offering up to 500 licenses to VMware View customers at no additional cost, and offering to new customers a 70% discount on Microsoft VDI Standard Suite subscription license and a 50% discount on Citrix XenDesktop VDI Edition annual license ($28 per device for up to 250 devices for one year).

Update: Brian Madden just published a video of RemoteFX in action.

Labels: Citrix, Microsoft, VDI

The popular security firm Core Security yesterday disclosed a serious security vulnerability found in all Microsoft hosted virtualization products, including Virtual Server 2005, Virtual PC 2007 (with and without SP1) and Windows 7 Virtual PC.

While Core Security is using the “hypervisor” terminology, this bug doesn’t affect any bare-metal virtualization platform Microsoft has, including Hyper-V and Hyper-V R2.

The vulnerability affects the virtual machine monitor (VMM) memory management.
It makes memory pages mapped above the 2GB available with read or read/write access to user-space programs running in a Guest operating system. By leveraging this vulnerability it is possible to bypass security mechanisms of the operating system such as Data Execution Prevention (DEP), Safe Structured Error Handling (SafeSEH) and Address Space Layout Randomization (ASLR) designed to prevent exploitation of security bugs in applications running on Windows operation systems.

Core Security even released a proof of concept code, vpdumper, that can be used to demonstrate the vulnerability.

The security vendor as well as worldwide press highlighted how this vulnerability affects Virtual PC platforms, but the most important aspect of this issue seems that it affects Virtual Server.
Over the years in fact virtualization.info collected multiple reports, from several countries, of small businesses that are running Virtual Server 2005 in production, still today, using it to virtualize mission critical applications like databases and mail servers (it doesn’t matter if Microsoft officially supports the scenarios or not).

Those companies may risk much more than single users running stand-alone Virtual PC and Windows 7 Virtual PC to execute spare VMs from time to time.

Core Security informed Microsoft about this vulnerability in August 2009.
It took four months to confirm the vulnerability and involve the product team and other relevant groups.
It took another five months to confirm that the vulnerability doesn’t affect Hyper-V.
At the end Microsoft informed Core Security about its plan to consider mitigating the issue in a future release of the affected products. Which means that at today there’s no fix for this.

Labels: Microsoft, Security

For a long time Microsoft decided to not enter the VDI market, leaving the competition with VMware to its trusted partner Citrix.
Over time, the strategy slightly changed: in January 2008 the company acquired the startup Calista, and in July 2009 it released its first connection broker as part of Windows Server 2008 R2.

Considering its position in the industry, virtualization.info speculated that Microsoft is just waiting for the right moment to enter the VDI space in a serious way.

Now several parts reports that this week Microsoft will make its move, by lowering the VDI cost with a more friendly license and by making VDI more powerful, with the merge of Calista technologies in the RDP code.

The announcement may come March 18 at 9am PST, according to this new website that recently surfaced online. 

Update: With a short note on its corporate blogs, Citrix announced that it will join Microsoft in this announcement.
This should clarifies once again that the two are not going to compete anytime soon despite the overlapping offering.

Labels: Calista, Microsoft, VDI

Microsoft has a free hypervisor: Hyper-V.
Citrix too (and it’s open source now): XenServer.

Microsoft has a virtualization management console: System Center Virtual Machine Manager (SCVMM).
Citrix too: Essentials for XenServer.

Microsoft has a VDI connection broker:  Remote Desktop Connection Broker.
Citrix too: Desktop Delivery Controller (DDC, part of XenDesktop 4).

Microsoft has an application and streaming virtualization platform: App-V (acquired by Softricity in 2006).
Citrix too: Application Isolation Environment (AIE, part of XenApp 6).

The presentation virtualization technology (below called “session virtualization”) at the core of XenApp is the only component that strictly depends on Microsoft platform.

Now, rather than competing head to head with Microsoft with such broad product portfolio, unlikely as long as XenApp represents the most significant part of its revenue, Citrix’s strategy is to support Microsoft overlapping technologies in a seamless way and add some features on top (even if this may cause some friction in the sales channel).

The company is doing so with Essentials, which is available for Hyper-V since early 2009 and it’s even free for SMBs. And it’s replicating with XenApp now.

With a provoking article Citrix just highlighted how the new XenApp 6.0 platform introduces tight integration with Microsoft App-V, in three areas:

Publish App-V packages directly from XenApp as "Dual-mode" applications
Through our new App-V integration, Microsoft App-V sequences can be published using the same workflows and wizards as all other applications managed through XenApp. Admins can leverage native XenApp Application Virtualization and Session Virtualization configuration parameters and policies to make Microsoft App-V sequences available for online and offline use.  Dual-mode fallback is also provided, enabling end-points to access App-V applications from a consistent interface, even if the device is incapable of running the application locally, for example when a user needs to access the application from a Mac PowerBook or iPhone.

Manage App-V client plug-in using Citrix Receiver
The Microsoft Application Virtualization Desktop Client can now be managed and delivered as a plug-in for Citrix Receiver. With XenApp, App-V sequences can now be delivered to lightly-managed endpoint scenarios, even if the end-point is not a member of the Microsoft Active Directory domain.  As a result, new App-V use cases and access scenarios are enabled, including the delivery of applications to a consultant or for companies with "Bring your own computer" (BYOC) initiatives.

Subscribe to App-V packages using Citrix Dazzle
With Citrix Receiver, users gain self-service access to applications through an enterprise app storefront. Admins can advertise App-V packages with all other XenApp published applications and services for easy, on-demand access by users.

Labels: Citrix, Microsoft

VMware’s activity on competition is intensifying in this period of the year: after last month’s comments on Citrix HDX performance and Essentials for Hyper-V sales volume, as well as the cost of managing Hyper-V, they focuses on Microsoft hypervisor’s stability.

Eric Gray, Senior Engineer on the VMware Competitive Team, writes on the topic on its personal blog vCritical, suggesting that ESX has a critical advantage over Hyper-V (and Xen and KVM) because Microsoft relies on 3rd party general-purpose drivers while VMware offers hardened, stress-tested drivers — ready for your toughest enterprise workloads.

Gray mentions a presentation performed by the Microsoft Technical Fellow Mark Russinovich, who confirmed that Windows crashes largely depend (70%) by 3rd party driver code (while Microsoft code is responsible only 5% of times).

The benefits vs shortcomings of relying on generic Windows drivers for Hyper-V is a old theme and while this specific article doesn’t add anything new to the discussion (despite it quotes a couple of real-world examples) the comments offer something new.

Ben Armstrong, Program Manager on Core Virtualization at Microsoft, jumps in and offers a different perspective on drivers quality and how it impacts hypervisors stability. He suggests that Microsoft drivers receive more thorough testing and validation than VMware drivers do:

…what is missing here is some scope. Yes, 70% of crashes are caused by drivers – without understanding the % of systems crashing compared to systems deployed this number is accademic.

Think of it this way (leaving out company names):

Company 1: Of 10,000 deployments 1,000 crashed in our code and 700 crashed in the drivers.

Company 2: Of 10,000 deplouments 300 crashed in our code and 700 crashed in the drivers.

Company 1 says: ~40% of crashes happen in the driver code on our system.
Company 2 says: ~70% of crashes happen in the driver code on our system.

Given all of this data – we can actually see that the driver quality is the same (in this hypothetical scenario). Unfortunately in this case we do not have the full numbers on how many deployments VMware / Microsoft has, how many crashes there are total, etc…

The purpose of this presentation (when originally given) was to talk about the health of the Windows ecosystem and platform as a whole. This is why the figure is a comparision of driver crashes to total crashes.

To make an assertion about the quality of drivers on one platform versus another platform – which you need is the percentage of driver crashes compared to total number of deployments.

Neither Microsoft or VMware have provided this information…

The entire comments thread is worth a read.

Labels: Microsoft, VMware

Microsoft recently launched a new, free TechNet Virtual Lab where customers can install and try Citrix XenDesktop 4.0 with Windows Server 2008 R2 with Hyper-V and System Center Virtual Machine Manager (SCVMM) 2008 R2.

Users will work through the steps to simulate deploying multiple virtual desktops, and will first capture a reference image using the Citrix Provisioning Server for Desktops. Next, the user will verify that the computer can be booted from a diskless client computer. The user will create multiple virtual machines using the reference computer as a template. Finally, the user will use the Citrix Desktop Delivery Controller and System Center Virtual Machine Manager to create a group of virtual desktops and deploy them to end devices. At the end of this lab, the user will have worked through all the steps required to implement Citrix XenDesktop using Microsoft System Center Virtual Machine Manager and Microsoft Hyper-V Server.

Set up an enterprise lab just to try XenDesktop may be expensive and certainly time consuming. So this is highly recommended to anyone interested in the Citrix VDI platform.

Labels: Citrix, Microsoft, Training

In December 2009 Microsoft acquired the run-book automation firm Opalis Software.
At that time the company anticipated that Opalis technology would be integrated in the System Center product family and that it would become the automation layer for Hyper-V and Azure virtualization.

Today Microsoft offers additional details about when the integration with happen: integrations packs for UNIX, Red Hat RHEL and Novell SLES Linux will be released in Q2 2010, while integration packs for Service Manager 2010, Configuration Manager (SCCM) 2007 R2, Virtual Machine Manager (SCVMM) 2008 R2 and Data Protection Manager (DPM) 2010 will appear in Q3 2010.

A tighter integration anyway can’t be expected before 2011 when Microsoft plans to release the next major version of Opalis.

The automation platform has been included in the Microsoft Server Management Suite Enterprise (SMSE) and Datacenter (SMSD) licenses which will cost more starting July 1. 

Labels: Microsoft, Opalis Software, Platform Orchestration

Microsoft just published the beta version of a new Infrastructure Planning and Design guide.
Titled Dynamic Data Center, this 43-pages blueprint on what Microsoft defines “a combination of automation, control, and resource management software with a well-defined topology of virtualization, servers, storage, and networking hardware”.

The guide is divided in five main parts:

• Determine the Dynamic Data Center Scope
  This part helps to define the scope and determine the workloads that will be included in the Dynamic Data Center project. It’s divided in three steps:
• Determine the proposed initial workloads for the Dynamic Data Center.
• Select the workload fault-tolerance approach (including Load Balancing, VM-Level and Host-Level Clustering, as well as Application-Level Fault Tolerance).
• Determine the initial size of the Dynamic Data Center.
  
• Design the Virtualization Hosts
  This part helps to design hosts that meet the capacity, performance, placement, and fault-tolerance requirements of the organization. It’s divided in three steps:
• Group the workloads
• Design the host’s Hardware configurations
• Determine host connectivity requirements
• Design the Software Infrastructure
  This part helps to design the software infrastructure in a way it can provide directory and authentication services, virtual machine management, configuration management, software distribution/inventory/patch management, event monitoring and collection, remote desktop services and hardware management.
• Design the Dynamic Data Center Storage
  This part helps to design the storage infrastructure in a way it meets requirements of capacity, performance delivery, fault tolerance and manageability. It’s divided in four steps:
• Design the storage system
• Design the host storage connections
• Design the storage switches
• Select the backup approach
• Design the Network Infrastructure
  This part helps to design the network infrastructure to accommodate the requirements accumulated throughout previous steps. It’s divided in three steps:
• Design network switches
• Design the hardware load balancers
• Design the firewalls

What the guide doesn’t cover instead is:

• Utility service providers or hosting companies offering cloud computing, cloud computing platforms, cloud platform services, or cloud infrastructure services.
• Windows Live™ network of Internet services or other consumer-oriented, cloud services.
• Microsoft Online Services or other third-party software-as-a-service organizations delivering Microsoft infrastructure technologies.
• VM guest planning is a complex topic, and although it informs the scoping in this guide, it is considered a related but separate exercise from planning the Dynamic Data Center infrastructure.

Now, while this document is a good starting point to plan and design what Microsoft calls a dynamic data center, it’s nowhere near the level of completeness that a customer need to design his infrastructure from scratch. And yet, it’s a great way to envision how massive is the investment that any company has to dedicate to data center design.

After reading this document, any customer would recognize some value in the emerging unified/fabric computing trend that companies like Egenera and Cisco are leading.

Labels: Microsoft, Papers

New month, new rebuttals in virtualization-land.
Evidently, virtualization players still consider the marketing skirmish very helpful to increase sales (virtualization.info has a slightly different opinion) so this March we have VMware leading three major campaigns against competitors.

Two of them are defensive, one is not:

• VMware View PCoIP vs Citrix XenDesktop HDX
• Volume of Citrix Essentials for Microsoft Hyper-V sales
• Cost of managing Microsoft Hyper-V vs VMware vSphere

PCoIP vs Citrix XenDesktop HDX

At the beginning of February Citrix sponsored a competitive analysis performed by Miercom.
The 7-pages report compares protocol performance of Citrix XenDesktop 4 (with ICA/HDX) and VMware View 4 (with PCoIP) and these are the conclusions:

In a comparison of Virtual Desktop Infrastructure (VDI) implementations, Citrix XenDesktop 4 provided better overall performance when compared to VMware View 4

XenDesktop 4 used 64% less bandwidth than View 4 with PCoIP for typical tasks

Flash video was delivered with an average of 65% less CPU usage, 89% less bandwidth, and excellent Quality of Experience by XenDesktop 4 compared to View 4

Overall, XenDesktop 4 uses system resources more efficiently and is capable of scaling more effectively

VMware answered last week, informing that they were not contacted by Miercom and that they have no insight about how test were conducted.
Of course VMware offered its point of view on each point,

At this point customers just have to decide which company has the nicest logo and which guy has the brightest smile to believe to one set of claims over the other. 
Luckily, Brian Madden jumps in and provide an impartial, long, detailed analysis that definitively is worth a read.

Volume of Citrix Essentials for Microsoft Hyper-V sales

At the beginning of March, a blog independently run by VMware employees published interesting speculations about the sales volume of Citrix Essentials for Hyper-V.

The article, written by Michael Hong, Senior Product Marketing Engineer, suggest that Citrix so far sold a very, very low number of Essentials for Hyper-V, because he apparently was the first one to recognize a major bug in the Workflow Studio setup.

Workflow Studio is part of the Essentials suite and the issue Hong encountered prevents its installation, but the Citrix support didn’t solve the issue and closed Hong’s support ticket without reasons.

Hong also notes that Citrix doesn’t have more than a bunch of posts on the its support forum about Essentials for Hyper-V.

For sure readers can’t wait to hear what Citrix has to answer on this…

Cost of managing Microsoft Hyper-V vs VMware vSphere

This is an old classic.
At the beginning of March VMware decided to cover a cost comparison table that Microsoft recently published.

The table compares several vSphere editions against a System Center bundle called System Center Management Suite Datacenter (SMSD), showing how the Microsoft way is significantly less expensive than VMware offering (at least half the price):

Of course there are a number of issues in the comparison that VMware pointed out.

In some cases VMware is completely right in highlighting how Microsoft doesn’t detail enough the difference between implementations of the same feature (vSMP support for example).
In other cases VMware wants Microsoft to drop comparison between some features because they are too different (VMware DRS vs Microsoft PRO for example) but here’s the a lot of room to debate.

While totally misleading, the sense of those marks is more like “We have this feature. The customer can use it in some way”.
Is it possible to pretend that a simple comparison table like this one (or the ones that VMware produces) offers an insightful qualitative analysis of implementation cost for each listed feature?
The customers that are looking for such in-depth side-by-side analysis aren’t going to research more on their own? Are they supposed to make their purchase decision just looking at this chart?

This is way the whole “my product is better than yours” marketing effort is a complete waste of time.

Labels: Citrix, Microsoft, VMware

Almost seven months after the launch of a public beta program, Microsoft finally released App-V 4.6 as part of the Desktop Optimization Pack (MDOP) 2010.

The new application virtualization platform introduces support for Windows Server 2008 R2 and Windows 7 (including 64bit editions), as well as the upcoming Office 2010.

The last point is particularly important because the new virtualized version of Office 2010, dubbed Click-To-Run, depends on this version of App-V.

App-V 4.6 also introduces a new feature called Read-Only Cache Mode, which is particularly useful in VDI environments. Ruben Spruijt offered a detailed analysis of the feature:

App-V 4.6 now supports the ability to configure the App-V client to use the package cache as a read-only file. This capability was added to support multiple App-V client machines access to a single package cache this functionality is very useful in Server Hosted Virtual Desktops (VDI) and Remote Desktop Services (Terminal Services) scenarios. With VDI there is a big impact on (central) storage. The IO and capacity increases while using vDesktops. Storage needs much attention while designing and maintaining a VDI environment. Microsoft App-V shared cache functionality will decrease the storage capacity impact.  The App-V package cache normally exists inside of a file with .FSD filename extension.

The FSD file must be placed in a location that can be accessed by the vDesktops; the location of the package cache should have read-only access and should perform well.  Direct Attached Storage (DAS) or central (SAN) storage will meet the storage criteria…

Microsoft updated its Infrastructure Planning and Design Guide to include information about App-V 4.6 new features.

Labels: Microsoft, Releases

Exactly one year ago two well-known virtualization experts Ruben Spruijt (Solution Architect and CTO at PQR) and Jeroen van de Kamp (Enterprise Architect and CTO at Login Consultants) released an independent, non-sponsored performance analysis comparing ESX 3.5, XenServer 5.0 and Hyper-V 2008.

The benchmark, specifically designed to measure desktop virtualization workloads (served by Terminal Services and VDI platforms), was so valid that Citrix decided to embrace the Virtual Reality Check methodology to measure XenDesktop 4 performance.

Twelve months later the two are back with a new comparison. This time they put side by side Citrix XenServer 5.5, Microsoft Windows Server 2008 R2 Hyper-V and VMware vSphere 4.0 Update 1, comparing them against their new workload simulator Virtual Session Indexer (VSI) 2.0.

The most interesting thing is that all tests were performed on HP hardware equipped with the new Intel Xeon 5500 Series CPUs (codename Nehalem), and compared to Virtual Reality Check 1.0 results obtained on previous generation processors.
Performance are almost doubled with both XenServer and vSphere, and with Hyper-V R2, performance are up 154%.

Once again, if you are involved in a desktop virtualization project this performance analysis is a mandatory reading.

Labels: Benchmarks, Citrix, Microsoft, Papers, VDI, VMware

This Tuesday Microsoft released its first security bulletin for Hyper-V.

The vulnerability, which could lead to a Denial of Service (DoS), consists in a malformed sequence of instructions that a locally authenticated user has execute on a guest operating system. 
It applies to all versions of Hyper-V, including Windows Server Hyper-V (2008, 2008 R2 and Server Core editions) and the stand-alone Hyper-V Server (2008, 2008 SP2 and 2008 R2 editions).

The vulnerability doesn’t affect the guest operating systems, it cannot be remotely exploited, it cannot be exploited by anonymous users inside the guest OS and it can’t be used for privileges escalation inside the hypervisor layer.

Additional information about affected files available here.

Labels: Microsoft, Security

At the end of January Microsoft silently updated its Linux Integrated Components package to version 2.0, introducing the long awaited support for Red Hat Enterprise Linux (RHEL) guest operating systems in Hyper-V.

Microsoft announced future support for Red Hat operating systems in July 2009, since the open source vendor joined the Server Virtualization Validation Program (SVVP).
Customers had to wait no less than seven months to finally have a version of Hyper-V Linux Integrated Components that supports RHEL 5 (including 5.2, 5.3 and 5.4 versions, both 32 and 64bit).

Like for Novell SUSE Linux, Microsoft doesn’t include in the package the optimized drivers for mouse. To have those customers need to rely on Citrix, which is offering them as open source through the Project Satori.
On top of that Linux Integrated Components still only supports Linux virtual machines with a single virtual CPU.

In July 2009 Microsoft also released the package as open source, and despite the drama behind that launch, the move should guarantee that every major Linux distribution will be included over time.
Let’s just hope that the process will not take seven months for every distribution out there.

Thanks to HyperVoria for the news.

Labels: Microsoft, Red Hat

At the beginning of January Microsoft launched its Platform-as-a-Service (PaaS) cloud computing offering: Windows Azure.
Despite the company’s Chief Architect Ray Ozzie said that Azure will be able to compete with Amazon EC2 and similar Infrastructure-as-a-Service (IaaS) clouds, this component is not yet accessible, or at least we couldn’t find it, and Microsoft didn’t even officially confirm it exists.

A couple of months ago virtualization.info suggested that the IaaS component of Azure may appear in March, because Microsoft is going to release a cloud toolkit that month. 
It seems that Azure will indeed start hosting virtual machines in March 2010 according to TechTarget:

…Microsoft has announced plans to add support for Remote Desktops and virtual machines (VMs) to Windows Azure, and the company also says that prices for Azure, now a baseline $0.12 per hour, will be subject to change every so often.

Prashant Ketkar, marketing director for Azure, said that the service would be adding Remote Desktop capabilities as soon as possible, as well as the ability to load and run virtual machine images directly on the platform. Ketkar did not give a date for the new features, but said they were the two most requested items…

The quoted part of the article doesn’t mention the timeframe that was originally published but somebody has been fast enough to quote it:

Microsoft is expected to add support for Remote Desktops and virtual machines (VMs) to Windows Azure by the end of March, and the company also says that prices for Azure, now a baseline $0.12 per hour, will be subject to change every so often…

March or not, Microsoft has acknowledged once again that Windows Azure will host virtual machines. This will put the company in direct competition with Amazon and VMware on the public/private cloud front.

If Microsoft can host a significant number of companies on Azure, it may improve the market perception around Hyper-V. Additionally, customers that will be able to experience and judge the Microsoft approach to cloud computing without running expensive pilots, may build confidence in a Hyper-V powered private cloud.

It’s likely that VMware will answer by launching its project Redwood and showing what happened to those $20 million invested in Terremark.

Labels: Cloud Computing, Microsoft

The capability to overcommit memory is an evident advantage that VMware has over competition. 
Marketing departments, partners, and even customers engage endless debates about its value and usefulness in every scenario. 

Microsoft has been particularly vocal in downplaying the importance of memory overcommitment, even if its President of Server and Tools Division, Bob Muglia, candidly admitted: 

we definitely need to put that in our product.

Maybe just a few people remind that Microsoft actually equipped its hypervisor with memory overcommitment capability and that was ready to appear in Hyper-V 2008 R2.
For some reasons anyway the company pulled the feature during the beta phase without disclosing when customers would finally get it.

The time is about to come, apparently.

After the launch of Windows 7, of course Microsoft is already working on Windows Server 2008 R2 post-RTM builds and Internet abounds of leaked screenshots about those.
One of them, published by Softpedia and related to the build 7700.winmain.100122-1900, reveals that Microsoft has reintroduced Dynamic Memory controls in the Hyper-V control panel:

This doesn’t really mean that the feature will appear in the next Windows Server 2008 R2 service pack or in Windows Server codename 8, but its presence in this build confirms that Microsoft is getting ready to reintroduce it.

At this point it will be hard to argue with VMware on the value of memory overcommitment (albeit it still makes sense to believe that it’s not useful in every scenario). 

Labels: Microsoft

Tracking the virtualization industry for more than six years (virtualization.info was launched in September 2003) has been a challenging, time-consuming and sometimes tiring task. But there always are fun moments.

The best ones come from the never ending skirmish between VMware and Microsoft marketing departments (and their allies), that in turn highlight the negative aspects of the competitor.
The VMware solution is too expensive and doesn’t manage anything but the virtual infrastructure, says Microsoft.
The Microsoft solution is not mature enough, it’s full of bugs (because it comes from Windows) and has hidden costs too, says VMware.

The effect that this exercise has on customers has been brilliantly summarized by Scott Adams in one of his recent Dilbert strips:

Reporting about new “evidence” that proves or disproves each others’ points doesn’t add much value to the discussion and virtualization.info tries to avoid doing so as much as possible.

Anyway, let’s make an exception this time, because it’s 2010 since 30 days and we didn’t get any pointless debate yet for the new year. And because it’s worth showing that old habits die hard even when it’s clear that they provide no results.

At the end of 2009 Information Week published an article titled 9 Reasons Enterprises Shouldn’t Switch to Hyper-V. Written by Elias Khnaser, Practice Manager, Virtualization & Cloud Computing at Artemis Technology, it contains a lot of criticism about Hyper-V memory management, security, live migration, maturity, cost and more.
Microsoft answered back in early January with Setting the Record Straight - 9 Reasons Why Hyper-V is a Great Choice for Enterprises, where Christopher Steffen, Principal Technical Architect at Kroll Factual Data, answered each point.
Of course this answer provoked a reaction, which will provoke another reaction and so on, forever.

Now, there’s a big chance that even after reading both articles (and subsequent reactions) you won’t change your opinion about both products if you already have one, or you are more confused than ever if you didn’t have one.
The reason is not just that for each claim there’s a counter-claim that balances the previous. It is also that such claims are completely ineffective to win the trust of a reader.

Everybody knows in fact that a person usually knows the product that he uses the most better. Unless you work with ESX and Hyper-V on daily basis, while they serve the same workloads in the same environment, and you have the same level of training to operate both platforms, your perception may be different about the two.
On top of that, everybody knows that people have the natural inclination to prefer something over something else. And this depends on a certain background or a previous experience that compromised the perception of a product or its manufacturer.
Last but not least everybody knows that it’s a common practice to secure testimonials by trading some unrelated benefits, like volume discounts, free licenses, extended support, public exposure, and more with the customer. That doesn’t mean that every case study is faked, and this certainly is not a reference to Microsoft and Kroll Factual Data. It just means that readers have no way to know what claim is genuine and what has been “encouraged,” thus they cannot trust anything.

The only real way to verify which product is the best one for a company is to run pilots and compare solutions on real-world duties. It is expensive and it is time consuming, but it definitively provides more concrete information than the debate above.

Labels: Microsoft

As most readers know, Microsoft offers three Hyper-V editions: the one coming with the full version of Windows Server 2008 R2, the one included in a stripped down version of Windows called “Server Core”, and the stand-alone Hyper-V Server.

The Server Core edition of Windows lacks the well-known GUI, has a limited .NET support and many other OS components are completely missing. It’s done so to reduce the OS surface attack, mimicking minimal Linux distribution that are popular among security professionals.

Problem is that this turns the local management of Hyper-V hosts into a real nightmare. Pretty much everything has be done on the command line but Microsoft doesn’t provide a CLI interface to perform every task.

For example, if you want to unbind a certain network protocol from the host NICs, and this is a typical hardening best practice, there’s no way to do so. Until today.
Microsoft just published a previously internal tool called NVSPBind that can be used to list binded protocols, enable and disable any of them from the Hyper-V NICs.

C:\>nvspbind -d {F93672D9-9085-4EEF-9669154AD4391ED7} ms_server
Hyper-V Network VSP Bind Application 6.1.7672.0.
Copyright (c) Microsoft Corporation. All rights reserved.
acquiring write lock...success

Adapters:
{F93672D9-9085-4EEF-9669154AD4391ED7}
"pci\ven_8086&dev_10c9&subsys_a03c8086"
"Intel(R) Gigabit ET Dual Port Server Adapter":
    unbinding ms_server from Intel(R) Gigabit ET Dual Port Server Adapter
    unbinding ms_server from Intel(R) Gigabit ET Dual Port Server Adapter
    unbinding ms_server from Intel(R) Gigabit ET Dual Port Server Adapter
    unbinding ms_server from Intel(R) Gigabit ET Dual Port Server Adapter
    unbinding ms_server from Intel(R) Gigabit ET Dual Port Server Adapter
    unbinding ms_server from Intel(R) Gigabit ET Dual Port Server Adapter
applying changes...
cleaning up...releasing write lock...success
finished
C:\>

Labels: Microsoft, Security, Tools

An area where Microsoft doesn’t seem particularly active is the so-called virtual desktop infrastructure (VDI).
So far the company made just a few progresses, letting its partner Citrix dominate the scene and compete head to head with VMware.

Rather than on products, Microsoft is focusing on VDI licensing.
In July 2009 it introduced two new VDI licenses, the Microsoft Virtual Desktop Infrastructure Standard Suite and the Microsoft Virtual Desktop Infrastructure Premium Suite, on top of its well-known Virtual Enterprise Centralized Desktop (VECD).
Now the company may perform additional adjustments to its offering.

A couple of days ago TechTarget reported that Microsoft plans to modify the VECD to reduce the cost per user ($23/seat/year if you are a Software Assurance customer, $110/seat/year if you are not).

Microsoft doesn’t plan to abandon its per-seat model but will introduces changes to extend use rights, allowing device roaming.

Labels: Microsoft, VDI

Despite its huge potential, it’s pretty evident that the market is not embracing the application virtualization approach (to not be confused with presentation or desktop virtualization) anytime soon.

All the biggest vendors in the IT industry invested in application virtualization: Microsoft acquired Softricity in May 2006, VMware acquired Thinstall in January 2008, Symantec acquired Altiris in January 2007 and AppStream in April 2008, Novell distributes XenoCode with an OEM agreement since September 2008, and Citrix has its own engine as part of XenApp since a long time.

Regardless of this massive commitment, the top players above spent almost zero effort to push for application virtualization adoption.
The startups that were not acquired in the last three years are struggling to make any impact. See AppZero (formerly Trigence), Ceedo or Trustware.

Microsoft, which owns a large part of the application ecosystem, and can deeply influence the rest of it, doesn’t seem to have any interest in winning the race, even if it owns what was considered one of the best application virtualization engine in 2006: SoftGrid (now App-V).
This year we are going to see a virtualized and stream version of Office 2010, which is good start but nowhere near the kind of effort required to facilitate a mass adoption.

Or the industry is still too busy pushing for the adoption of hardware virtualization and its related applications (VDI, IaaS cloud computing), or the application virtualization technology is not mature enough to be useful outside specific niches, or simply there’s no need for application virtualization, and all the companies above just went deadly wrong with their investments.

On top of these options there’s another one: customers are looking for alternatives to application virtualization that are perceived as more flexible.
One of them may be the so-called offline VDI, powered by client hypervisors that expected later this year.

Labels: AppZero, Ceedo, Citrix, Endeavors Technologies, Microsoft, Novell, Symantec, Trustware, VMware, XenoCode

Along with System Center Data Protection Manager (SCDPM) 2010 and Visual Studio Team System (VSTS) 2010 Lab Management, Microsoft is about to release another product that will have extended virtualization capabilities: System Center Essentials (SCE) 2010.

Essentials (to not be confused with Citrix Essentials, which indeed is another management suite for Hyper-V) is a bundle of multiple System Center products, integrated by a single management console, that supports up to 50 servers and 500 clients.

The current version, SCE 2007, merges together System Center Operation Manager (SCOM) 2007 and Windows Server Update Services (WSUS) 3.0.
It doesn’t have any capability to manage Hyper-V because it doesn’t include the System Center Virtual Machine Manager (SCVMM) engine.

The upcoming version instead, now in Release Candidate status, will include the SCVMM 2008 R2 engine, which means that customers will be able to create and operate virtual machines, perform physical to virtual (P2V) and virtual to virtual (V2V) migrations, manipulate templates and snapshots, and more.

It’s not clear if SCE 2010 will be able to manage VMware environments like SCVMM 2008 R2.

Microsoft published a 14-minutes demo here.

Labels: Microsoft, Platform Management

As most virtualization.info readers know by now, Microsoft is finally approaching the .NET developers with a virtualization-friendly edition of its upcoming IDE Visual Studio 2010.

The product will be called Visual Studio Team System 2010 Lab Management, and will integrate Hyper-V R2 and System Center Virtual Machine Manager (SCVMM) 2008 R2 to offer a virtual lab automation platform that competes against products like VMware Lab Manager, VMLogix LabManager, Surgient Virtual Automation Platform and others.

Microsoft took forever to leverage its huge MSDN community to let Hyper-V slip into new customers’ sites.
Ironically, the company is doing it right now that VMware, who rules the developers world thanks to Workstation, seems to have lost interest in it.

On top of the new Visual Studio 2010, which may be released in Q2 2010, Microsoft recently released another tool called VM Factory:

Visual Studio 2010 VM Factory is the reference implementation of a software solution that automates the creation of Visual Studio 2010 and Team Foundation Server 2010 virtualized environments. The purpose of this project is to build prescriptive guidance around virtualization of the Visual Studio 2010 and guidance for full automation of the creation of virtual machines using the VM Factory. The goal is to help users with the installation and configuration of virtualized environments with least effort and maximum automation.

The blueprint, which is released under the Creative Commons 3.0 license, includes the following pieces:

• Rangers Virtualization Guidance
  • Focused guidance on creating a Rangers base image manually and introduction of PowerShell scripts to automate many of the configuration tasks.
  • Virtualization guidance looking at the “why” and “how” to use virtualization for Team Foundation Server and Visual Studio, including planning, pre-requisite software, use of non-Microsoft virtualization technologies and introducing use case scenarios.
• Rangers Factory Package and Guidance
  • Reference walk-through documentation on how to install, configure and support a Microsoft internal or an external factory to automate the installation of Team Foundation Server and Visual Studio environments.
  • Microsoft Deployment Toolkit metadata and PowerShell scripts used to create a Rangers factory.

Labels: Microsoft, Virtual Lab Automation

One aspect where Microsoft is extremely weak in the virtualization realm is the VMs backup and restore.

The market doesn’t offer many choices to Hyper-V customers and even the Microsoft own enterprise disaster recovery solution, System Center Data Protection Manager (SCDPM) 2007, leaves much to desire.

Things are slowly changing. In part because existing partners like NetApp are more committed to release products that support Hyper-V now that Microsoft is gaining some concrete market share.
In part because the VMware strategy is pushing even its most loyal partners into Microsoft’s arms.
In part because Microsoft is working on a better integration between Hyper-V and most of its other products.

Data Protection Manager 2010, which is about to reach the Release Candidate status, is part of this ongoing effort.

virtualization.info already reported that the product will finally allow to protect virtual machines inside the Windows Server 2008 Cluster Shared Volumes (CSVs).

Given this new capability, it’s highly desirable to automatically protect new virtual machines that are stored in a cluster.
Microsoft gives guidance on how to manipulate DPM 2010 with PowerShell scripts to do so.

Labels: Disaster Recovery, Microsoft

Yesterday Microsoft and HP announced a 3-year agreement to spend $250M on several fronts: Hyper-V and System Center, Windows Azure, Exchange, SQL Server and more.

The problem with such announcements (see the 3-year alliance with EMC for another example) is that just a few (to not say nobody) really understand what’s the difference between before and after the deal.
The language used in the press announcements never helps.

Microsoft and HP already are very good partners. Customers expect the option to have Microsoft products inside their brand new HP servers at the purchase time because this happens since a lot of year.
So this deal requires some clarifications (of course we’ll focus on Hyper-V, System Center and Azure):

• Part of the investment is to develop new integrated offerings that involve Hyper-V Server, System Center Essentials, HP Virtual SAN appliance (formerly Lefthand Networks), HP Flex Fabric and HP Operations Center.
  Specifically, HP is now able to OEM Hyper-V in future bundles rather than just offering it as a preconfigured option for its ProLiant customers. Additionally, HP can also OEM System Center Essentials and SQL Server.
• Part of the investment is to pump up the sales channel (at worldwide level) and the professional services that will sell and support the products above.
  The investment in the sales channels will be 10x the actual one.
• The announcements has been used to highlight that Microsoft will continue to buy HP ProLiant and BladeSystems for the Windows Azure cloud infrastructure.
  This doesn’t mean that HP is or will be the only hardware provider for Azure.

Now, despite Steve Ballmer and Mark Hurd clarified that this deal was discussed no less than two years ago and that it was approved in April 2009, it’s very likely that its execution has been impacted by the VMware-Cisco-EMC (VCE) alliance.

The leadership of HP is threatened by the entrance of Cisco in the server market primarily because Cisco is not just trying to sell bare metal, it’s trying to sell a complete hardware/software platform that doesn’t require any interaction with other vendors.
On paper, this is the peace of mind for every data center admin.

For now this platform is limited to a “simple” bundle of vSphere, the Unified Computing System (UCS) blade system and the Ionix Unified Infrastructure Manager, a super console kindly developed by EMC.
But soon the three may come out with a totally integrated implementation of the VMware vCloud APIs, turning a set of specifics into a concrete private cloud facility (Vblock 4 anyone?).

HP didn’t just lose the opportunity to do this first. It also (partially) lost a key partner, VMware, which is executing its cloud vision with someone else.
So HP buys 3Com to reinforce its enterprise-grade networking offering, and announces a major commitment to work with Microsoft. 
And the target is not just the VCE Vblock offering. The virtualization market’s money is on professional services. So the target also is the joint venture that Cisco and EMC created: Acadia.

At the lowest level, this agreement is not about developing some amazing new virtualization platform that can slam the Vblock. It is about avoiding that Cisco sells too many Vblocks.

Labels: Alliances, Cloud Computing, HP, Microsoft

The Microsoft Operations Framework (MOF) is a sort of blueprint that companies can use to standardize the way they manage their IT, pretty similar to what the Information Technology Infrastructure Library (ITIL) set of practices does.

Virtualization is such a complex technology that involves OSes, networks and storage administration, security hardening, performance monitoring and even applications troubleshooting in some cases. As we said a long time ago, virtualization professionals are sort of super heroes.
There’s a real need for operational frameworks to handle virtual infrastructures but unfortunately nor ITIL neither MOF are currently covering this aspect of the IT governance.

A step in the right direction may come from Microsoft, which released the first beta of what they call Reliability Workbook for Hyper-V, as part of the existing MOF 4.0 framework.
The Reliability Workbooks are task lists (in Excel format) that Microsoft recommends to follow to monitor and maintain the health and reliability of their products.

Before using a Reliability Workbook a company has to fine tune it, performing a list of steps:

Once done, the customer can move on the spreadsheet, which includes 5 pages besides the Overview and Acknowledgements:

• Monitoring Activities
• Maintenance Activities
• Health Risks
• Standard Changes

Here’s the Health Risks table:

It is still really really basic and honestly, considering the available resources and experience built with four generations of MOF, Microsoft could do so much more than this, showing the authority in system administration that its marketing department use to sell against VMware.

It’s the first beta anyway. Hopefully customers will have something better for the RTM.

Labels: Microsoft

Finally Microsoft launched Windows Azure, the Platform-as-a-Service (PaaS) cloud computing offering that competes with products like Google App Engine.

In Q3 2009 the company said that Azure will be more than just a PaaS cloud, hosting (Hyper-V) virtual machines pretty much like Amazon does with its Elastic Computing Cloud (EC2). Which makes it a hybrid IaaS + PaaS cloud. But Microsoft didn’t clarify if the IaaS offering would be launched at the beginning of 2010 with the PaaS one.

To find out virtualization.info explored a bit Azure.

The first thing that is worth to note is the pricing scheme:

As you can see, there’s an entry level package (Introductory Special) that’s available free of charge.

It comes with a small computing instance, a terminology that reminds the Amazon one to describe its virtual machines. 
A careful analysis of the Azure features included in each package confirms that Microsoft computing instances are compared to virtual servers. And like Amazon, the company offers three sizes to customers:

And like Amazon does, every customer is limited to maximum 20 small computing instances (or equivalent computing resources) by default. If you want more you need to ask for them.

After signing as new Microsoft customer and purchasing the free Introductory Special, we are required to wait for a confirmation that our account is ready to use the cloud. This took just 10 minutes in our case.
At this point we are able to access the Azure control panel to provision the services we need:

Here there’s nothing more than Storage Account and Hosted Services options. After selecting the the latter, we are invited to upload our application and configuration files:

Apparently, there’s no way to freely use the small computing instance that is included in the Introductory Special plan as a hosted virtual machine. 
Similarly, the Storage Account service above can’t be used to upload an existing virtual machine as Ray Ozzie suggested during his PDC 2009 keynote. Thus the IaaS offering is not yet enabled.

Labels: Cloud Computing, Microsoft

Last Friday Microsoft announced the acquisition of Opalis Software, a run-book automation company founded in 1998 in Canada with 70 employees (according to LinkedIn) and over 300 customers (according to Opalis).

Data center orchestration is one of the most important areas where virtualization will expand in the coming years, as soon as customers will realize that their virtual infrastructures are reaching such a scale and complexity to become inefficient.

VMware and Citrix already invested in this area.
VMware acquired the Swiss startup Dunes Technologies in September 2007, and it’s now offering their solution for free, as part of vSphere 4.0 platform, under the name of Orchestrator.
Citrix offers an orchestration framework called Workflow Studio since January 2009.

Microsoft isn’t shy to say that the Opalis Software technology will be integrated in its System Center portfolio and that it will serve as the automation layer for Hyper-V virtualization and Azure cloud computing.

With this acquisition Microsoft doesn’t need much more to compete end-to-end against VMware on virtualization. Yes, the company still doesn’t have a client hypervisor, and didn’t clarify if it plans to become a major VDI player on its own or not, but the technology to deliver both missing pieces is there.
The Microsoft problem is that its vision about virtualization is nowhere near the VMware’s one. They have a lot of products which could be integrated to form an impressive end-to-end offering, but the today’s reality is completely different. And of course the customers’ trust in the Microsoft capability to compete with VMware reflects this.

Update: In October, The 451 Group analysis firm reported that this acquisition is around $60M.
Opalis raised $25M in venture capital funding and scored over $10M in revenue.

Labels: Acquisitions, Microsoft, Opalis Software, Platform Orchestration

At the last Microsoft PDC conference, the company’s Chief Software Architect Ray Ozzie said that customers will be able to upload their (Hyper-V) virtual machines inside the soon-to-be-launched cloud computing platform Windows Azure.

More clearly, Ozzie said that Microsoft will support the hosting mode offered by Xen-based (like Amazon EC2) and VMware-based cloud architectures, but, besides that, the company didn’t provide any detail about the availability of this IaaS component or the way it works.

The only thing we know for sure is that Microsoft is preparing a toolkit to guide its customers to extend their virtual data center into the Azure cloud.
Now we also know that this toolkit is going to be available in March 2010.

The Platform-as-a-Service (PaaS) component of Azure will be launched January 1, 2010. Maybe Microsoft plans to announce an Infrastructure-as-a-Service (IaaS) part too, and that one will arrive in March.

Labels: Cloud Computing, Microsoft

After a couple of months in beta, Citrix releases Essentials 5.5 for Hyper-V just before the holidays.

This version of the management platform for the Microsoft hypervisor includes a new technology called StorageLink Site Recovery.

This feature allows the Hyper-V administrators to control the replication features that their SAN arrays without using multiple consoles. From the Essentials console they can test the recovery process with what-if analysis, and restore the protected VMs in isolated, test networks.

The notable thing is that StorageLink Site Recovery is available for every version of Essentials, including the Express one which is free of charge (but it won’t appear there before Dec. 23).
HP announced its support for this technology a long time ago and now confirms integration with StorageWorks SANs.

Citrix published a bunch of videos to show how it works here.

At the moment there are not many (virtualization-aware) solutions for disaster recovery of Hyper-V virtual machines and it’s not clear if Microsoft is going to release its own or not. So Citrix has a good chance here to be considered in most comparisons.

Labels: Citrix, Disaster Recovery, Microsoft

HP always offered a basic capacity planning tool to its customers that want to use ProLiant servers for virtualization.
In November 2005 it released one for Microsoft Virtual Server 2005. In March 2007 it released one for VMware VI 3.0.

Yesterday the company released also one for Microsoft Hyper-V R2. This one is not a web tool like the previous versions, but a 50MB Windows application that customers can download and use without restrictions.

To collect data from physical servers, the Sizer tool interacts with the Microsoft Assessment & Planning (MAP) Toolkit (both 3.x and 4.x are supported) or the Windows Performance Monitor, but it can also import information from other tools.
Once data is available, this tool produces a detailed Bill of Materials (BoM) that includes servers and storage equipment, with pricing specified for the customer’s country.

It also includes an update engine which automatically downloads new inventory parts and refreshes prices.
A check of this engine shows that the capacity planning engine is marked as version 4.0:

Labels: Capacity Planning, HP, Microsoft, Releases