elstir

Hi - by now you might have read there's 17,000 of us attending VMworld in San Francisco. Huge crowds, just as Rick Vanover predicted. Lots of energy and excitement as you can imagine. This post is designed to bring some of the show to you, assuming you're not attending and queuing up to a session 45 mins before the start.

The expo hall started Monday. The attendees who found us were entertained to see 'the biggest little booth' at VMworld. Here's a view. Mike Neil, our GM of Windows Server and Server Virtualization, filmed this 20-minute video from the VMworld blogger lounge (aka, The Cube). After the expo floor closed, we and Citrix hosted a Tweetup. Great conversations and crowd - not to mention the excellent vanilla bean beer made by Thirsty Bear. The discussions reflected the still maturing adoption of virtualization:

• a software gent from Fort Collins is attending VMworld to learn more about server consolidation. They have 100+ servers and are running out of room. We discussed what Hyper-V could do for him, and then after a bit of time it was revealed they're 99% SPARC shop with Java apps. We agreed that he's got to tackle the hardware before the Virt, otherwise Oracle OpenWorld is the show to attend ;-)
• three gents from Rockwell Collins. One gent works desktop support, and the other two gents work the datacenter. We had a good laugh about the power struggle between both camps during VDI projects, but not the case at Rockwell Collins to their credit. The most interesting part of the discussion was the one gent's ability to use session-based access to his email while on the airplane.
• I spoke with a few folks from Joann Fabrics. They're a big System Center shop, and are using ESXi w/ their Windows apps. I stepped into a funny trap when I told them, "My wife loves Joann Fabrics." I guess they hear that line all the time. Dana and crew were kind enough to stop by the booth yesterday, and hopefully we'll see you at a future MMS.

That evening, the Aug. 31 edition of USA Today starting hitting the streets starting in the East Coast. The front news section included this 'open letter' advertisement to VMware customers from Microsoft's Brad Anderson. It turned some heads so far. And, of course, VMware had an appropriate response.

Edwin Yuen published a blog worth reading, as it summarized our demos in the booth. Here's an excerpt:

So at this year's VMworld, we are demoing the cloud solution that Outback Steakhouse created using Windows Azure Platform. Working with a partner, Outback Steakhouse developed and deployed an online marketing campaign in less than eight weeks - the flexibility and scalability of the cloud allowed them to support overwhelming customer response.  The marketing campaign met its goal of 500,000 fans in only 18 days. It's a great example of IT being able to satisfy business and marketing demands with a fast, cost-effective solution.

We will also demo how we're helping customers use the same tools to control and manage Windows Azure-based applications, as they would applications running on Windows Server. Customers can use System Center Operations Manager to monitor the health of applications, whether the apps are on-premises or on Windows Azure, and in return get a complete view of how well all their IT services are running. We showed this demo at Microsoft Management Summit 2010. This solution provides the critical capability to manage your applications regardless of the infrastructure they may run on, whether it be your datacenter or the public cloud with Windows Azure.

Yesterday's keynote was probably the best of the 7 VMworld conferences that I've attended. It was a combination of game day celebration, painting pictures of the future, and showing what here or coming. There were several holes in the presentation. Go here to watch/listen to Mike Neil, Simon Crosby (CTO at Citrix) and Harry Labana (CTO at Citrix) comment on the keynote.

One of the more entertaining lines, or at least the takeaway, is when CEO Paul Maritz said the OS is no longer the center of innovation. His point is that the OS isn't going away, but rather the future innovation will be in virtualization, app frameworks and end-user access. This statement supports his company's lofty P/E ratio and investments in future revenue streams such as SpringSource, vBlock, View 4.5. Thankfully, we offer all that and more today:

OS (Windows Server Hyper-V), app framework (.NET), cloud-scale OS (Windows Azure), common identify and mngt (AD, System Center), desktop optimization (App-V, RDS, RemoteFX).

The meetings yesterday convinced me that Windows Azure much different than EC2 (off-premises IaaS) and VMware's vCloud (private cloud, IaaS), but there's little understanding of what it can do for people today. Coca-Cola, The Tribune Company, RiskMetrics, and Outback Steakhouse are examples that help people understand.

I hope you found this recap useful.

Patrick

 

I’ll be posting more this week about what you’ll see when you visit the Microsoft booth (#1431) at VMworld 2010 U.S. conference next week. You can read an overview at NetworkWorld.

But last week a virtualization industry insider and consultant who has followed Microsoft virtualization for 5+ years and has met with Microsoft at the last 5-6 VMworld conferences, wrote the following to me:

“I really don't get why you guys insist on exhibiting at VMworld. The number of people stopping by must be huge to justify it.”

His email made me pause. If this gentleman doesn’t know why we’re there, then many others must wonder the same. I/we answered that question lots of times the first few years at VMworld, but the questions have diminished in the last few years. So I reflected on why we participate at VMworld, and wanted to share our thinking. At the high level, VMworld attendees are Microsoft customers: 

·        Approximately 75% of VMworld attendees run Windows Server (estimate based on industry analyst data of virtualized OSes)

·        About 50% of VMworld attendees run Microsoft server applications, such as Exchange, SQL, Sharepoint (estimate based on market share data)

·        Nearly 90% of VMworld attendees run Windows XP/Vista/Windows 7 and MS Office (estimate based on market share data)

We first chose to exhibit at VMworld in 2005 because it was marketed as an industry event, and we wanted feedback on our virtualization solution from the early adopters. Since then, most booth visitors at VMworld have encouraged Microsoft’s participation at the event because they benefit from industry competition and they have questions for us. Here’s a summary of highlights from each year:

·        At VMworld 2005, attendees kicked the tires on Virtual Server, and Diane Greene stopped by the booth to thank us for being there. This was the only time a VMware executive came to our booth.

·        At VMworld 2006, the booth was so crowded we gave out 1,500 t-shirts in 150 minutes on Day 1; we handed out 4,600 t-shirts during the entire show. We generated 300+ sales leads. See a recap here.

·        At VMworld 2007, we had 200+ sales leads and a busy booth with people interested in “Viridian” and “SoftGrid.” See blog post here.

·        At VMworld 2008 Europe, we launched Windows Server 2008 and gave out 2,000 copies of the launch kit. It was the first time European customers saw us at VMworld, and we received over 200 sales leads. See Mike Neil’s interview here.

·        At VMworld 2008, we reached 3,800 attendees in less than 90 minutes, which translated into 175K visits in 7 days to a marketing website. We also co-presented with VMware about the Server Virtualization Validation Program, and met with representatives of 55 VMware partners.

·        At VMworld 2009 Europe, over 250 people attended our session. We handed out 500 copies of eval software, and met with representatives of 30 VMware partners.

·        At VMworld 2009, the first year of our 10x10 booth, we attracted 2,500 new followers to Microsoft via social media, and generated about 100 sales leads.

·        At VMworld 2010 and VMworld Europe 2010 … we shall see.

We get a ton of value from participating in the industry discussions at VMworld and, based on the feedback we get at the booth, conference attendees appreciate that we’re there.

I’m interested in your thoughts on our attendance at VMworld. Should we try to expand our presence at VMworld, or some other industry event? Should we opt out and focus on adding more virtualization content to Microsoft conferences (e.g., MS Management Summit, MS TechEd, Worldwide Partner conference)? Or should we just print lot, lots more Microsoft t-shirts?

Patrick

The savvy editors of NetworkWorld chose 3 weeks before VMworld 2010 conference to publish an online tech debate between Microsoft and VMware. The topic: who has the better virtualization platform. The editorial summary reads:

VMware is the kingpin of virtualization, but the game is changing fast and Microsoft is baking the technology into the very core of many products. Which company has the best approach?

Read the debate here. Following is an excerpt:

There was a perception among early adopters of server virtualization that Microsoft didn't have a rich feature set. That's not the case. More than a year ago we further simplified and expanded clustering nodes, and added live migration for zero-downtime migrations of virtual machines between Hyper-V servers.

Hyper-V also provides high availability with transparent and automatic failover of virtual machines. With service pack 1 of Windows Server 2008 R2, we're adding Dynamic Memory and a new high-fidelity remote desktop protocol.

Lastly, you should read Enterprise Strategy Group's lab results that show Hyper-V performance versus physical devices, with 95% to 99% of the performance of physical disks, and 89% to 98% of performance of the tested workloads compared to what can be achieved on physical machines.

At Microsoft we believe virtualization is so critical we've made it part of our server OS, our management tools and our cloud strategy. As a result, VMware is missing critical features: the ability to manage both physical and virtual machines; the ability to get information about the application running within the virtual machine located on-premises or cloud; the ability to manage virtual machines from Microsoft, VMware and soon Citrix.

Go vote (for Microsoft) and make yourself heard by leaving a comment.

If you're attending VMworld in San Francisco, stop by the Microsoft booth (#1431). It'll be easy to miss us in the 10x10 booth ;-), so check back here and our Twitter feed for more details on booth demos.

Patrick

We are really excited to announce the availability of the Hyper-V Linux Integration Services for Linux Version 2.1. This release marks yet another milestone in providing a comprehensive virtualization platform to our customers. Customers who have a heterogeneous operating system environment desire their virtualization platform to provide support for all operating systems that they have in their datacenters. We have supported Linux as a guest operating system on our virtualization platform from the days of Virtual Server and continue to enhance our support in that regard.

The following features are included in the 2.1 release:

Driver support for synthetic devices: Linux Integration Services supports the synthetic network controller and the synthetic storage controller that were developed specifically for Hyper-V.

Fastpath Boot Support for Hyper-V: Boot devices take advantage of the block Virtualization Service Client (VSC) to provide enhanced performance.

Timesync: The clock inside the virtual machine will remain synchronized with the clock on the host.

Integrated Shutdown: Virtual machines running Linux can be gracefully shut down from either Hyper-V Manager or System Center Virtual Machine Manager.

Symmetric Multi-Processing (SMP) Support: Supported Linux distributions can use up to 4 virtual processors (VP) per virtual machine.

Heartbeat: Allows the host to detect whether the guest is running and responsive.

Pluggable Time Source: A pluggable clock source module is included to provide a more accurate time source to the guest.

                                                                

This version of the integration services for Hyper-V supports Novell SUSE Linux Enterprise Server 10 SP3, SUSE Linux Enterprise Server 11, and Red Hat Enterprise Linux 5.2 / 5.3 / 5.4 / 5.5.

Customers can obtain the Linux IC’s via the Microsoft Download Center at this link: http://www.microsoft.com/downloads/details.aspx?displaylang=en&FamilyID=eee39325-898b-4522-9b4c-f4b5b9b64551

 With the release of Beta of Service Pack 1 for Windows Server 2008 R2 a number of you have asked about Service Pack 1 for the standalone Microsoft Hyper-V Server 2008 R2, and whether the new capabilities of Dynamic Memory and RemoteFX will be available for it. Absolutely, both Dynamic Memory and RemoteFX have been developed for Microsoft Hyper-V Server 2008 R2 as well.

 

In order to get these capabilities for the Microsoft Hyper-V Server 2008 R2, you will need to install the Beta of Service Pack 1 on Microsoft Hyper-V Server 2008 R2. Note that the first wave of the Service pack installer is only in 5 languages (English, French, German, Japanese and Spanish), so if you try and apply the package to Microsoft Hyper-V Server 2008 R2 (which has 11 language packs installed by default) you will rightly see the following screen

 

 

 

 It’s pretty simple to uninstall these language packs to thereafter install the Service pack. In order to uninstall the language packs, there is nifty utility included (lpksetup.exe). Launch this from an administrator’s command prompt and select “Uninstall display languages”.

On the next screen, select the all languages other than the five (English, French, German, Japanese and Spanish). Of course if you want to save some additional disk space, you can uninstall other languages as well and leave just the language that you use in your environment, Click next and let the tool do its job. Thereafter you can apply Service Pack 1. Enjoy!!

Vijay Tewari

Principal Program Manager, Windows Server Virtualization

Back in February we started working with partners and TAP customers on a toolkit to create private cloud environment.  See our February blog about the Dynamic Infrastructure Toolkit for System Center.

A couple updates were announced Monday at our Worldwide Partner Conference. First, the tool has a new name: System Center Virtual Machine Manager Self-Service Portal 2.0 (the portal). Second, we reached the release candidate milestone. You can download it here. A release candidate milestone means it’s feature complete, and is nearing completion. We expect the portal to be finalized and released to the web in calendar Q4.

You can read more about the self-service portal tool over at the System Center team blog (here). Here's an excerpt:

The self-service portal provides the following features that are exposed through a web-based user interface:

• Configuration and allocation of datacenter resources: Store management and configuration information related to compute, network and storage resources as assets in the VMMSSP database.
• Customization of virtual machine actions: Provide a simple web-based interface to extend the default virtual machine actions; for example, you can add scripts that interact with Storage Area Networks for rapid deployment of virtual machines.
• Business unit on-boarding: Standardized forms and a simple workflow for registering and approving or rejecting business units to enroll in the portal.
• Infrastructure request and change management: Standardized forms and human-driven workflow that results in reducing the time needed to provision infrastructures in your environment.
• Self-Service provisioning: Supports bulk creation of virtual machines on provisioned infrastructure through the web-based interface.Helps business units to manage their virtual machines based on delegated roles

Patrick

======================================================

Preamble: The point of this series, and the spirit in which it is written, is to take a holistic approach at the issues facing our customers, discuss the complexities with regard to memory management and explain why we’re taking the approach we are with Hyper-V Dynamic Memory. This isn’t meant to criticize anyone or technology, rather to have an open and transparent discussion about the problem space.

======================================================

In the past few blogs we’ve covered Page Sharing and Second Level Paging. Today, let’s dig into what we’re delivering with Hyper-V Dynamic Memory in Windows Server 2008 R2 SP1 as well as our free hypervisor Microsoft Hyper-V Server 2008 R2 SP1. So what is Dynamic Memory?

Dynamic memory is an enhancement to Hyper-V R2 which pools all the memory available on a physical host and dynamically distributes it to virtual machines running on that host as necessary. That means based on changes in workload, virtual machines will be able to receive new memory allocations without a service interruption through Dynamic Memory Balancing. In short, Dynamic Memory is exactly what it’s named.

Let’s dive in an explain how all this works starting with the new Dynamic Memory settings. Here are the new settings available on a per virtual machine basis. Here’s a screenshot:

Dynamic Memory In Depth

With Hyper-V (V1 & R2), memory is statically assigned to a virtual machine. Meaning you assign memory to a virtual machine and when that virtual machine is turned on, Hyper-V allocates and provides that memory to the virtual machine. That memory is held while the virtual machine is running or paused. When the virtual machine is saved or shut down, that memory is released. Below is a screenshot for assigning memory to a virtual machine in Hyper-V V1/R2:

With Hyper-V Dynamic Memory there are two values: Startup RAM and Maximum RAM and it looks like this:

Startup RAM is the initial/startup amount of memory assigned to a virtual machine. When a virtual machine is started this is the amount of memory the virtual machine will be allocated. In this example, the virtual machine will start with 1 GB.

The Maximum RAM setting is the maximum amount of memory that the guest operating system can grow to, up to 64 GB of memory (provided the guest OS supports that much memory). Based on the settings above, here’s an example of what the memory allocation could look like over a workday...

As you can see, the workload is dynamically allocated memory based on demand.

Next, let’s look at the Memory Buffer.

Memory Buffer: In one of the earlier blogs posts in this series, we discussed the complexity of capacity planning in terms of memory. To summarize, there is no “one size fits all” answer for every workload as deployments can vary based on scale and performance requirements. However, one consistent bit of feedback was that customers always felt more comfortable by providing additional memory headroom ‘just in case.’

We completely agree.

The point being you want to avoid a situation where a workload needs memory and Hyper-V has to start looking for it. You want some set aside memory as buffer for these situations, especially for bursty workloads.

The Dynamic Memory buffer property specifies the amount of memory available in a virtual machine for file cache purposes (e.g. SuperFetch) or as free memory. The range of values are from 5 to 95. A target memory buffer is specified in percentages of free memory and is based on current runtime memory usage. A target memory buffer percentage of 20% means that in a VM where 1 GB is used, 250 MB will be ‘free’ (or available) ideally for a total amount of 1.25 GB in the virtual machine. By default, Hyper-V Dynamic Memory uses a default buffer allocation of 20%. If you find this percentage is too conservative or not conservative enough, you can adjust this setting on the fly while the virtual machine is running without downtime.

This takes us to the last Dynamic Memory setting, Memory Priority.

Memory Priority: By default, all virtual machines are created equal in terms of memory prioritization. However, it’s very likely you’ll want to prioritize memory allocation based on workload. For example, I can see a scenario where one would give domain controllers greater memory priority than a departmental print server. Memory Priority is a per virtual machine setting which indicates the relative priority of the virtual machine's memory needs measured against the needs of other virtual machines. The default is set to ‘medium’. If you find that you need to adjust this setting, you can adjust this setting on the fly while the virtual machine is running without downtime.

 

Dynamic Memory Works Over Time With A Few VMs…

I’ve explained the per VM settings and shown how this would work with a single virtual machine, but how does Dynamic Memory work with multiple virtual machines? Below is an example to show just how Dynamic Memory works. I’ve kept this example simple on purpose to avoid confusion. Let’s assume I have a small server with 8 GB of memory. I’m going to run three virtual machines, one from Finance, Sales and Engineering. Each virtual machine is given the same settings: Startup RAM = 1 GB and Maximum RAM = 4 GB. With these settings, each virtual machine will start 1 GB and can grow up to 4 GB as needed.

Virtual Machine Start. On the left graphic below, you can see three virtual machines starting. Each virtual machine is consuming 1 GB of memory for Startup RAM. On the right graphic below, you can see the total amount of memory being used in the entire system ~3 GB.

15 minutes later. The Finance VM is running reports while the Engineering VM starts an analysis job. With Dynamic Memory, the Finance VM is allocated 3 GB of memory, the Engineering VM is allocated 2 GB of memory while the Sales VM remains at 1 GB. System wide, the server is now using 6 GB of its 8 GB or 75% of the total physical memory.

30 minutes later. The Finance VM is running reports while the Engineering VM starts an analysis job. With Dynamic Memory, the Finance VM is allocated 2 GB of memory, the Engineering VM is allocated 3.5 GB of memory while the Sales VM remains at 1 GB and a fourth VM, Service VM is started using 1 GB of memory. System wide, the server is now using 7.5 GB of its 8 GB of memory for VMs. At this point the server is fully allocated in terms of memory and is using its memory most efficiently.

 

At this point, the question I’m always asked is, “What now? What if a virtual machine still needs more memory? Does the parent start paging?”

No.

At this point, Dynamic Memory will attempt to reclaim pages from other virtual machines. However, in the absolute worst case where no free pages are available, the guest operating system will page as needed, not the parent. This is important because the guest operating system knows best what memory should and shouldn’t be paged. (I covered this back in Part 5...) Finally, when free memory does become available from other virtual machines, Dynamic Memory will move memory as needed.

Over-Subscription & the CPU Analogy

One argument we routinely hear is that there’s nothing wrong with over-subscription. Customers tell us that they take a bunch of physical servers, virtualize them and run the server with over-subscribed CPUs without issue, so why is this an issue with memory?

Great analogy, wrong conclusion.

Example 1: Suppose you are running 8 physical servers at 10% utilization, virtualize them and run those 8 virtual machines on a single server for a total of ~85% utilization. In this example, you’re not over-subscribing the CPU and the server still has 15% CPU headroom.

Over-subscription is this…

Example 2: Suppose you are running 8 physical servers at 50% utilization, virtualize them and run those 8 virtual machines on a single server. The single server would max out at 100% utilization, but because the workloads require ~400% utilization, performance would be terrible. What would you do? Move virtual machines to other servers of course to avoid over-subscription. In short, what you really want to do is maximize resource utilization to get the best balance of resources and performance.

That’s exactly what we’re doing with Hyper-V Dynamic Memory.

Customer Requirements & Dynamic memory

When it comes to virtualization and memory, virtualization users have repeatedly provided the following requirements:

1. Use physical memory as efficiently and dynamically as possible with minimal performance impact. Customers investing in virtualization hosts are purchasing systems with larger memory configurations (32 GB, 64 GB, 128 GB and more) and want to fully utilize this system asset. At the same time, they’re purchasing this memory to provide superior performance and to avoid paging.
2. Provide consistent performance and scalability. One frequent comment from virtualization users is that they don’t want a feature with a performance cliff or inconsistent, variable performance. That’s makes it more difficult to manage and increases TCO.

You got it. Here’s why we’ve chosen the path we have with Dynamic Memory.

1. Dynamic Memory is truly a dynamic solution. Memory is allocated to virtual machines on the fly without service interruption based on policy.
2. Dynamic Memory avoids significant performance penalties by not adding additional levels of paging which can significantly impact performance
3. Dynamic Memory takes advantage of Large Memory Pages and is, in fact, optimized for Large Memory Pages
4. Dynamic Memory is a great solution for virtualizing servers and desktops (BTW, Dynamic Memory works fine with SuperFetch)

Cheers,

Jeff Woolsey

Principal Group Program Manager

Windows Server & Cloud, Virtualization

P.S. Here are the links to all of the posts in this blog series:

• Part 1: http://blogs.technet.com/virtualization/archive/2010/03/18/dynamic-memory-coming-to-hyper-v.aspx
• Part 2: http://blogs.technet.com/virtualization/archive/2010/03/25/dynamic-memory-coming-to-hyper-v-part-2.aspx
• Part 3: http://blogs.technet.com/virtualization/archive/2010/04/07/dynamic-memory-coming-to-hyper-v-part-3.aspx
• Part 4: http://blogs.technet.com/b/virtualization/archive/2010/04/21/dynamic-memory-coming-to-hyper-v-part-4.aspx
• Part 5: http://blogs.technet.com/b/virtualization/archive/2010/05/20/dynamic-memory-coming-to-hyper-v-part-5.aspx

Over at the Windows Server Division blog, Oliver says the beta is available to download. Oliver focuses on Windows Server 2008 R2 sp1, but Windows 7 sp1 beta is also available. All of this was announced today at our Worldwide Partner Conference in Washington, D.C.

Here's an excerpt from Oliver's post:

The two most important developments in SP1 for Windows Server 2008 R2 are:

Dynamic memory lets Hyper-V administrators pool available memory on a physical host and dynamically distribute it to any virtual machine(s) running on that host. So as the workloads on that physical workload change, requiring more or less memory, Dynamic Memory will let administrators change the memory allocation to their VMs without service interruption. For a deeper look at Dynamic Memory check here.

RemoteFX enhances Microsoft desktop virtualization. RemoteFX lets Windows Server 2008 R2 administrators provide an even richer and user-transparent desktop virtualization experience. RemoteFX delivers rich content, independent of any graphics stack, to server-hosted virtual and session-based desktops, allowing them to support any screen content, including full-motion video, portable graphics stacks such as Silverlight, and 3D applications. Because it can use virtualized graphics on the server and advanced codecs , RemoteFX can deliver those experiences to a much wider array of target devices, including standard desktops and laptops but also an emerging slew of thin clients. You'll also be able to forward the USB ports of the local client to the virtual machine being accessed on the device - just like you can forward the local printer over RDP today. [UPDATE - see Max's new blog post on partner support for RemoteFX]

Make sure you check out the new SP1 Beta Resource page on Microsoft.com as well as the TechNet SP1 page -- and don't forget to grab the download here. 

Patrick

At HP's Tech Forum conference, HP started to market a VDI reference architecture, and deployment offer, with us and Citrix. The brochure can be seen here. The VDI offer can be seen here.

The nuts and bolts of the architecture includes:

• HP ProLiant BL460c server blades
• HP StorageWorks P4800 SAN
• HP Insight Control for System Center
• Citrix XenDesktop 4.0 enterprise edition
• Hyper-V R2

There are new HP infrastructure services for VDI to help you get started. HP's Dave Donatelli spoke of this VDI architecture (and more) toward the end of his keynote today. It starts around 1:08:00 of Donatelli's keynote.

Patrick

One of the fun parts of being in a team that develops Hyper-V is seeing it in action. Here at TechEd 2010 in New Orleans, Hyper-V and the System Center Suite are running the Hands-On-Lab infrastructure. Powering the labs for the 11,000 people here at TechEd are 25 x HP380G6’s. The detailed specs of each server is

·        HP DL 380G6

·        128 GB RAM

·        1.0 TB HDD

·        2 x Intel Xeon E5504, 2 Ghz

The attendees are accessing labs (running as virtual machines) from 350 workstations. There are a total of 192 labs running and on an average each lab takes just over 2 virtual machines. By the first day and a half of the conference, just over 1800 labs were launched leading to 3750 virtual machines being created and destroyed!!. By the way HOL11, Web Development in Microsoft Visual Studio 2010 is the most popular lab so far.

The infrastructure workloads, all running virtual machines, includes SQL Server 2008 SP1 for usage and performance statistics, Domain controller and DHCP for the lab domain, Web server for publishing the labs, System Center Operations Manager 2007 R2 for monitoring the health of the infrastructure, System Center Configuration Manager 2007 R2 for compliance monitoring and running scripted actions and System Center Data Protection Manager for backing up SQL and Hyper-V configurations.

I’d encourage folks who are here at TechEd to stop by this mini datacenter in Ballroom C and chat with folks there to get more details on the setup. Below are a few pictures of the setup. Enjoy!!

That's all it takes these days:).

System Center Operations Manager, reporting at work

Vijay Tewari

Principal Program Manager, Windows Server Virtualization

Every now and again, we like to use this forum to provide commentary and context on happenings in the industry. Today’s news from Waltham, Mass. and Palo Alto, Calif. is a good example. On the one hand, there’s a Microsoft competitor. On the other hand, there’s a Microsoft partner (see today’s joint webcast). And if there’s such a thing as a third hand, there are implications for Red Hat, perhaps even IBM.

So what’s our take?

First, the Microsoft and Novell alliance has been very successful. It’s been 3.5 years since the big announcement with Novell, and we have more than 475 customers who’ve chosen to work with Microsoft and Novell. We’ll continue to work with Novell around technical support and interoperability. The joint interoperability lab in Cambridge is very active, and we recently announced work together in high-performance computing. Oh, and did I mention today’s joint webcast on heterogeneous datacenter ;-)

Second, the vFolks in Palo Alto are further isolating themselves within the industry. Microsoft’s interop efforts have provided more choice and flexibility for customers, including our work with Novell. We’re seeing VMWare go down an alternate path. As one of many examples of our work with open source communities, we’re adding functionality to the Linux Integration Services for Hyper-V. In fact, we have an RC version of the Linux Integration Services, which support Linux virtual machines with up to 4 virtual CPUs. In fact, we’ll talk more of this on June 25 at Red Hat Summit. For more meat, see Matthew’s post about what you need to know about Linux on Hyper-V. And if you’re talking about technical support, let’s not forget the Server Virtualization Validation Program, which includes Red Hat, VMWare, Novell, Cisco and others.

Third, looks like VMWare finally determined that virtualization is a server OS feature. I’m sure we’ve said that once or twice over the years ;-). The vFolks now plan to ship a full version of a server OS with vSphere, and support it, to fulfill their application development and application deployment plans.

Fourth, this is a bad deal for customers as they’re getting locked into an inflexible offer. Check out the terms and conditions. For example: “Customers may run SLES with the accompanying patches and updates subscription entitled by a VMware purchase only in virtual machines running on VMWare vSphere 4.0 and 4.1 hosts that have active vSphere SnS with VMware.” So be sure not to drop support or you’ll invalidate your license. Or maybe just stick with small deployments.

Last, the vFolks have no public cloud offering, like Windows Azure, like Amazon EC2. While we’re demoing and building capabilities so customers have a common and flexible application and management model across on-premises and cloud computing, they’re stitching together virtual appliances to fill the void. Don’t forget – the next version of System Center Virtual Machine Manager will configure VMs from VMWare and Citrix.

What are your thoughts?

Patrick O’Rourke, director of communications, Server and Tools Business

Wow, 11,000 attendees, press, analysts and speakers here at TechEd. That’s a testimony to the passion, and excitement that this generates amongst our customers and partners.

 

Day 1 got off to a flying start with Bob Muglia and his inimitable style.  For those who did not get a chance to view the keynote you can catch the recording at

http://www.msteched.com/2010/NorthAmerica/Keynote01. Some really cool demos of System Center Virtual Machine Manager vNext and how we are taking our learning’s from running our datacenters and providing that technology to you (@ the 16 minute mark).

 

I was at the booth for almost the entire day and the traffic was great. We had lots of customers who came over and had great detailed technical conversations on their deployments with Hyper-V and SCVMM. Quote of the day “I love Hyper-V and am moving over 1800-2000 servers running VMware”.

 

There were a few questions from customers on network failover and load balancing. I want to take this opportunity to point out that our partners such as

Intel and Broadcom have powerful solutions for NIC teaming.

 

The bummer was when we lost Internet connectivity for a while at the convention center. Thankfully it was restored before the end of the day and we could return to a sense of digital comfort and connectivity.

 

Day 2 has a plethora of sessions on Virtualization and Management. Below is what we have in store for you today. Today’s is the big day for details on Dynamic Memory and RemoteFX.

Tuesday, Jun 08, 2010

8.00 - 9.15 am

WSV305: Deploying, Virtualizing, and Managing Linux and UNIX with Hyper-V

VIR318: Datacenter Virtualization: It's More than a Hypervisor

 

 

Tuesday, Jun 08 2010 9.45 – 11 am

WSV307: Windows Server 2008 R2 SP1

VIR06-Int (Interactive Session: Failover Clustering with Hyper-V Unleashed with Windows Server 2008 R2

VIR305: Microsoft RemoteFX: Rich Windows Desktop Experience for VDI and Session Virtualization

 

Tuesday, Jun 08, 2010 1.30 – 2.45 pm

VIR304: Hyper-V and Dynamic Memory in Depth

VIR311: Planning and Deploying Microsoft VDI with Management Technologies

 

 

Tuesday, Jun 08, 2010 3.15 – 4.30 pm

VIR307: Getting the Most out of Windows Server 2008 R2 Hyper-V with the Integration of the System Center Suite

 

 

 

Tuesday, Jun 08, 2010 5.00 – 6.15 pm

VIR310: Networking and Windows Server 2008 R2 Hyper-V: Deployment Considerations

 

 

 

 

Best schwag of the day, the brown Microsoft Hyper-V Server 2008 R2 teddy. Yes, I am biasedJ.

Come and see us at the Virtualization booths over on the left side of the expo hall under the RED Microsoft learning center.

 

Vijay Tewari

Principal Program Manager, Windows Server Virtualization

 

Hi, as I type this post while flying over the country, it just amazes me how far we have come from the days of original computing machines. There is something surreal in being able to blog from an aircraft at 37,000 feet.

I will be at TechEd in New Orleans starting this Monday Jun 7, 2010. We are really excited with the content that we have available for our customers. Besides the two big announcements on RemoteFX and Dynamic Memory there are a host of other sessions that we have for you. Below you will find a list of breakout sessions that are relevant to folks interested in delving into Server Virtualization and Management. Don’t forget that we also have some great hands-on labs and sessions on App-V, MED-V and a host of other technologies and solutions as well. Note that session times sometime change due to unavoidable circumstances so please check the latest schedules at the venue.

 

Date and Time

Session

Session

Session

Session

Monday, Jun 07, 2010

1.00 – 2.15 pm

VIR 206: Virtualization 360: Microsoft Virtualization Strategy, Products, and Solutions for the New Economy

VIR02-Int (Interactive Session): Hyper-V Live Migration over Distance. A Multi datacenter approach

 

 

Monday, Jun 07, 2010 2.45 – 4.00 pm

VIR317: Desktop Virtualization: You have a choice

 

 

 

Monday, Jun 07, 2010 4.30 – 5.45 pm

VIR204: Understanding how Microsoft Virtualization Compares with VMware

VIR07-Int (Interactive Session): Solving the VDI Licensing puzzle

 

 

Tuesday, Jun 08, 2010

8.00 - 9.15 am

WSV305: Deploying, Virtualizing, and Managing Linux and UNIX with Hyper-V

VIR318: Datacenter Virtualization: It's More than a Hypervisor

 

 

Tuesday, Jun 08 2010 9.45 – 11 am

WSV307: Windows Server 2008 R2 SP1

VIR06-Int (Interactive Session: Failover Clustering with Hyper-V Unleashed with Windows Server 2008 R2

VIR305: Microsoft RemoteFX: Rich Windows Desktop Experience for VDI and Session Virtualization

 

Tuesday, Jun 08, 2010 1.30 – 2.45 pm

VIR304: Hyper-V and Dynamic Memory in Depth

VIR311: Planning and Deploying Microsoft VDI with Management Technologies

 

 

Tuesday, Jun 08, 2010 3.15 – 4.30 pm

VIR307: Getting the Most out of Windows Server 2008 R2 Hyper-V with the Integration of the System Center Suite

 

 

 

Tuesday, Jun 08, 2010 5.00 – 6.15 pm

VIR310: Networking and Windows Server 2008 R2 Hyper-V: Deployment Considerations

 

 

 

Wednesday, Jun 09, 2010, 8.00 – 9.15am

VIR315: Windows Server 2008  R2 Hyper-V Performance Analysis: How You Can Get the Most out of Hyper-V

WSV309: Microsoft RemoteFX: USB and Device Support

 

 

Wednesday, Jun 09 2010 9.45 – 11 am

WSV316: Hyper-V and Storage: Maximizing Performance and Deployment Best Practices in Windows Server 2008 R2

VIR403: Virtualization FAQ, Tips, and Tricks

 

 

Wednesday, Jun 09, 2010 11.45 – 1.00 pm

VIR313: VDI from Microsoft and Citrix: What is it? How do I manage it? What benefits does it provide?

 

 

 

Wednesday, Jun 09, 2010, 1.30 – 2.45 pm

VIR03-Int (Interactive Session): Hyper-V for the VMware Administrator

 

 

 

Wednesday, Jun 09, 2010 3.15 – 4.30 pm

VIR316: Remote Desktop Session Host vs. Virtual Desktop Infrastructure Smackdown

 

 

 

Wednesday, Jun 09, 2010 5.00 – 6.15 pm

VIR09-Int (Interactive Session): Virtualization Round Table Discussion

 

 

 

Thursday, Jun 10, 2010 8.00 – 9.15 am

WSV13-Int (Interactive Session): Next Generation VDI with Microsoft RemoteFX

VIR314: Virtualization Scenarios for Business Critical Applications

VIR312: Realizing a Dynamic Datacenter Environment with Windows Server 2008 R2 Hyper-V and Partner Solutions

WSV315: Guest vs. Host Clustering: What, When, and Why

Thursday, Jun 10, 2010 9.45 – 11.00 am

VIR309: Microsoft System Center Virtual Machine Manager 2008 R2: Advanced Virtualization Management

 

 

 

Thursday, Jun 10, 2010, 1.30 – 2.45 pm

VIR01-Int (Interactive Session): Disaster Recovery and Virtualization Protection with Microsoft System Center Data Protection Manager (DPM) 2010

MGT312: Microsoft System Center Operations Manager and Virtual Machine Manager: Monitoring the Service Stack

 

 

Thursday, Jun 10, 2010 5.00 – 6.15 pm

VIR080-Int (Interactive Session): Virtualization: Security Tales from the Twilight Zone!

VIR322: Hyper-V Backup: A Look Under The Hood

 

 

 

Looking forward to seeing you all at TechEd.

Vijay Tewari

Principal Program Manager, Windows Server Virtualization

For those of you who think about desktop management strategies, and how virtualization might fit, I'd recommend reading Edwin Yuen's new blog post. Some of you may know Edwin as he's contributed to this blog, now blogs at Virtualization Planet, and has done lots of demos during keynotes.

Edwin's latest post describes our approach to desktop virtualization, the partners we work with on desktop virtualization solutions, a brief video with customers Expedia, Holland America and Group Health. Edwin's conclusion is:

 I would recommend working on developing your desktop virtualization strategy today.  This means researching and understanding what impacts virtualization can have on your desktops and what the advantages and disadvantages of each solution can be.  Click on the thumbnail below to explore the Microsoft Desktop Virtualization Hour Web site.  And if you are ready to get started with Desktop Virtualization or even if you are already deployed with Desktop Virtualization, you should take advantage of some special, limited-time offers from Microsoft to kick start your VDI implementation or replace your VMware VDI licenses.

If you're looking for something more meaty, such as planning and deploying VDI, then I recommend this on-demand webcast (registration required) by Michael Kleef. Here's the abstract:

In this webcast, we drill down into the architecture of the Virtual Desktop Infrastructure (VDI) and determine what key questions need to be asked and answered around required components, networking, capacity, and end-user experience. We also discuss how this can also fit within a Remote Desktop Services deployment.

 

Patrick

Hi, I'm Adrian Simays, Hyper-V technical architect at EMC.

EMC and Microsoft have been thinking big lately.  How big?  We’ve set out to build and test one of the largest Hyper-V environments in the world. 

And why not?  With the release of Windows Server 2008 R2, Microsoft introduced features such as clustered shared volumes (CSVs) for storing multiple VHDs on a single LUN, more efficient processing of networking traffic, as well as improved memory management and support for 256 logical processors.  Meanwhile EMC’s storage arrays are known for delivering unprecedented levels of scalability, availability and performance.  In particular, EMC’s Virtual Provisioning technology saves space and provides high levels of performance, which is something that solutions that tend to use CSVs can benefit from.  Combined, Windows Hyper-V and EMC storage platforms can easily support a large-scale, highly available virtualized configuration.  Plus, here at EMC, we continue to see wide adoption of Hyper-V including large scale deployments (this includes one of our customers who is currently running seven 16-node Hyper-V clusters on one EMC Symmetrix VMAX!).

So we set out to build a large scale Hyper-V cluster to capture the performance benefits of these new features. The goal of this test is to demonstrate how the environment scales using features made available with Windows Server 2008 R2 and EMC’s Symmetrix VMAX Enterprise Storage Platform.  Objectives for this test were fairly aggressive due to the size and scale of the configuration. We wanted to create 16 nodes in a Windows Failover Cluster with each parent node containing 64 child virtual machines for a total of 1024 virtual machines.  We set out to achieve 100,000 total IOPS and document the hardware layout as well as capture best practices.

So what have we found so far?  Well we found that we could exceed our total I/O goal with just 4 parent nodes containing a total of 256 VMs.  By using Live Migration we spread VMs across the parent nodes while leaving the physical data on the CSV volume.  This verified that the underlying storage platform, when appropriately configured can scale as we provide parallel access from multiple nodes in the Active/Active configuration provided with CSV.  We are continuing to collect performance data as we increase the number of parent servers and VMs in the environment.

Another key finding included some efficiencies in deploying an environment of this size.  We found that scripting the deployment of VMs using a SCVMM template was extremely easy and took about 15 minutes to deploy 5 VMs over the network (using two 1Gbit network links).  This is great for most environments but for 1024 VMs this would take 52 hours to complete.  By using EMC’s TimeFinder technology and using snapshots, the entire processing time could be significantly reduced to just a few minutes.  A volume was provisioned with Windows images that were configured via SYSPREP.  These images (aka templates) also included required applications.  By creating snapshot copies of the master volume and then importing these replicas as CSVs and finally instantiating the VMs we could reduce the time to create the VMs to a matter of minutes for each parent server.

Interested to learn more?  Come join us at Microsoft TechEd 2010 where EMC will present on this solution and the key findings during one of the Breakout Sessions (VIR207: Advanced Storage Infrastructure Best Practices to Enable Ultimate Hyper-V Scalability) as well as a joint whitepaper to be published shortly.  For more information, please visit: www.emc.com/microsoftvirtualization.

 Adrian

Hi, this is Bryon Surace again.  I’m a senior program manager on the Windows virtualization team at Microsoft. 

 

I wanted to draw your attention to the new Hyper-V Component Architecture Poster.   

The poster is a great visual tool to help in the understanding of the key features and components of the Hyper-V in Windows Server 2008 R2.  It highlights key Hyper-V components including:

§  Architecture

§  Virtual Networking

§  Virtual Machine Snapshots

§  Live Migration

§  Storage Interfaces

§  Storage Types

§  Storage Location and Paths

§  Import and Export

 

This large-format poster provides practical visual depictions of the Windows Hypervisor, live migration process, cluster shared volumes architecture, VMQ data paths, disk storage I/O path, and much more.

 

Download, Print, and Enjoy!

Bryon Surace

 

Hi, I’m Friea Berg from NetApp’s Virtualization Solutions team. Today, I’d like to share how Microsoft and NetApp have been helping customers get the most out of their virtualized infrastructure. Microsoft is a leader in applications, Windows, and management while NetApp leads in highly efficient and flexible storage. By integrating our solutions and joining forces, we provide the virtualized infrastructure building blocks that enable customer success. 

There are significant challenges and capabilities required at each layer of an end-to-end virtualization solution.  Our engineers have worked with Microsoft engineers for more than four years as Microsoft developed first Hyper-V and then R2, and we’ve built on our expertise with Microsoft technologies to deliver products tightly integrated with Hyper-V, System Center Operations Manager, System Center Virtual Machine Manager, Microsoft dynamic datacenter toolkits and Microsoft enterprise applications.  

As a result, NetApp and Microsoft have developed a unique virtualization solution that spans the datacenter from the desktop to the cloud. By combining Microsoft’s built in hypervisor and integrated System Center management platform with unique NetApp efficiency technologies such as primary storage deduplication and incremental only backup copies, our customers can maximize the utilization of both servers and storage to dramatically reduce power, cooling and datacenter space. Using Microsoft management tools you can monitor and manage an entire complex datacenter – including NetApp storage. 

 

The strength of our joint solution led Microsoft to name NetApp it’s Storage Solutions Partner of the Year and to establish a 3-year strategic alliance with virtualization as a major area of focus. Microsoft VDI partners Citrix and Quest have integrated directly into NetApp to leverage our native storage array-based cloning features to enable mass provisioning of desktop VMs, and Cisco recently published an Exchange 2010 on Hyper-V with NetApp validated design. Just this week, Microsoft and NetApp were named one of the top 10 cloud computing collaborations to watch. 

We believe so strongly in the benefits of our joint virtualization solution that in early 2009 NetApp was the first partner to guarantee that joint Microsoft and NetApp virtualization customers will use 50% less storage. Today, our customers and partners are the biggest advocates of our joint virtualization solution and have realized substantial cost-savings, improved SLAs, accelerated deployments and more. Below are a few examples: 

• Law firm Miller & Martin is headquartered in a 1917 building and had completely maxed out the power grid. By combining Windows Server 2008 R2 Hyper-V with NetApp efficiency technologies such as deduplication, the firm was able to cut its physical footprint 66% and reclaim 50% of used storage. Most importantly, physical space savings plus reducing power consumption by a third enabled the firm to avoid having to seriously pursue a costly datacenter expansion project. (details)  

• In Texas, one of the fastest growing cities in the country replaced 40 servers with just 4 and cut capex 54%. Deploying Microsoft virtualization with NetApp storage didn’t just enable City of Frisco to significantly cut costs but also to dramatically speed provisioning of new applications and facilitate interagency information-sharing to support groundbreaking initiatives. Where provisioning servers for a new project used to take 6-8 weeks working through procurement, it’s now done in minutes. (details)  

• In Austria, for the Vienna Chamber of Labor virtualizing SharePoint 2010 with Hyper-V and NetApp storage has yielded a very flexible, scalable infrastructure with substantial efficiency and data protection benefits including the ability to restore even to an individual item. (details)  

• Channel partner DM Consultants recently blogged about an Exchange 2010 installation involving 750 mailboxes, heavy usage profile, and 3 Mailbox Databases of approximately 150GB each, stored on a NetApp iSCSI solution and running without any issues. “Stunning.” (details)  

• Customers are now embarking on larger and more ambitious data center projects. Hosting provider nGenX, for example, has built a cloud computing offering that provides a single portal for customers to manage their virtual and physical environments in any of five data centers and developed an automated geo-diverse disaster recovery strategy. The company expects to add 1,000 VMs this year while reducing support costs by up to 40%. (details)  

Visit NetApp at booth 801 at TechEd June 7-10 or stop by the Microsoft virtualization kiosk on Tuesday, June 8th, to learn more about how NetApp and Microsoft are delivering on our promise to help customers transform their data centers to achieve greater efficiencies, reduce costs, and respond faster to changing business needs.

Friea Berg

This post is for all the Microsoft channel partners who read this blog. You'll want to check out our new virtualization partner profitability toolkit (here). David Greschler's blog here sums it up better than I could. Here's an excerpt:

Every partner knows about the great opportunity to sell virtualization technology and services. For a while, that meant one thing: working exclusively with VMware.

Times have changed.

Microsoft’s virtualization technology is more affordable and seen by many as technically comparable, and the company is gaining market share. Indeed, Microsoft Hyper-V continued its ascent in Q409, growing 215 percent year over year, according to the market research firm, IDC.

Nearly every VMware partner we talk to recognizes that Microsoft is a major player in virtualization and tells us they would like an opportunity to build a virtualization practice that includes both VMware and Microsoft. They recognize that having a dual practice allows them to become more trusted advisors to their customers. But they want a way to better understand the economic model.

To that end, Microsoft today is announcing a new virtualization partner profitability toolkit. The kit, available here on the Microsoft Partner Network, contains a profitability modeling tool that partners can use to see exactly how much adding a Microsoft practice can mean to their business. It’s basically an interactive spreadsheet, built with midsize business customers in mind, to help partners visualize the economic differences of doing project work around technology from one vendor versus being technology agnostic and building a virtualization practice that makes use of technology from multiple vendors. 

I just caught this article on the topic of channel partners choices for virtualization, and the toolkit. Here are some quotes:

"Very generally, if the customer has a really good view of what they want to see at the end of a project and can clearly articulate that, Microsoft has the best opportunity because its management wraps around its virtualization where VMware has separate [management] stacks," Russell said. 

"There will always be cases where if a customer already has investment in VMware, deploying [more VMware] might be less expensive [than Hyper-V]," said Jim Vanden Boom, a virtualization solutions manager at CDW LLC. But, again, said Vanden Boom, "since Hyper-V is included in the base Microsoft Server OS, if you're going to have to buy a new server OS anyway, Hyper-V could make more sense."

Kathi Grumke, VandenBoom's colleague and another solution specialist at CDW, concurred. "Customers are looking for agnostic analysis so they can make a decision of what makes sense so whatever tools we can deploy help us have that conversation with the customer," she said. "They may have invested in VMware already. Maybe it's coming up for renewal. Should they go for that or look at Hyper-V? The toolset gives us the opportunity to have that trusted adviser conversation." 

 While you check out the toolkit, don't forget that dynamic memory and RemoteFX are coming with SP1 of WS08 R2 (a no cost addition if you have R2).

Patrick

Hi readers,

Due to the planned MSDN/TechNet blog upgrade, we have to re-post blogs posted during the week of May 17.  We wanted you to know you'll be seeing 2-3 posts again. Thanks for your patience with us.

Patrick

=====================================================================

Preamble: The point of this series, and the spirit in which it is written, is to take a holistic approach at the issues facing our customers, discuss the complexities with regard to memory management and explain why we’re taking the approach we are with Hyper-V Dynamic Memory. This isn’t meant to criticize anyone or technology, rather to have an open and transparent discussion about the problem space.

=====================================================================

In my last blog, we covered some follow-up questions about Page Sharing. Today, we’ll discuss Second Level paging. To discuss the implications of using Second Level Paging, let’s put virtualization aside, take a step back and level set and start by discussing Virtual Memory and Paging.

Virtual Memory At A High Level

Modern operating systems employ virtual memory. Virtual memory is a way of extending the effective size of a computer’s memory by using a disk file (as swap space) to simulate additional memory space. The operating system keeps track of which memory addresses actually reside in memory and which ones must be brought in from disk when needed. Here are a few of the common memory management functions performed by modern operating systems:

• Allow multiple applications to coexist in the computer's physical memory (enforce isolation)
• Use virtual addressing to hide the management of physical memory from applications
• Extend the system's memory capacity via swapping

Virtual Memory In Depth

Let’s dive in deeper. For that, I’m going to reference a TechNet article that discusses the Windows Virtual Memory Manager. If you’d like to read the full article it is here: http://technet.microsoft.com/en-us/library/cc767886.aspx. A second article I highly recommend on virtual memory is this one from Mark Russinovich: http://blogs.technet.com/markrussinovich/archive/2008/11/17/3155406.aspx

From the TechNet article:

Sharing A Computer's Physical Memory

Operating systems that support multitasking allow code and data from multiple applications to exist in the computer's physical memory (random access memory) at the same time. It is the operating system's responsibility to ensure that physical memory is shared as efficiently as possible, and that no memory is wasted. As a result, an operating system's memory manager must contend with a problem called memory fragmentation. Memory fragmentation refers to the situation where free (available) memory becomes broken into small, scattered pieces that are not large enough to be used by applications. In the example shown here, free memory is separated into three separate blocks.

Once free physical memory becomes fragmented, an operating system can consolidate free memory into a single, contiguous block by moving code and data to new physical addresses. In this case, the three blocks of free memory were consolidated into one larger block by moving system memory upward and application 1 downward in physical memory.

If an application accesses its code or data using physical memory addresses, the application may encounter problems when the operating system moves its code and data. A mechanism must be provided for applications to access their code and data no matter where the operating system moves them in physical memory.

Virtualizing Access to Memory

A common solution is to provide applications with a logical representation of memory (often called virtual memory) that completely hides the operating system's management of physical memory. Virtual memory is an illusion that the operating system provides to simplify the application's view of memory. Applications treat virtual memory as though it were physical memory. Meanwhile, the operating system can move code and data in physical memory whenever necessary.

In a virtual memory system, the addresses applications use to access memory are virtual addresses , not physical memory addresses. Every time an application attempts to accesses memory using a virtual address, the operating system secretly translates the virtual address into the physical address where the associated code or data actually resides in physical memory. Because the translation of virtual addresses to physical addresses is performed by the operating system, applications have no knowledge of (or need to be concerned with) where their code and data actually reside.

Extending Virtual Memory Through Swapping

When applications access memory using virtual addresses, the operating system is responsible for translation of virtual addresses to physical addresses. As a result, the operating system has total control over where data and code are physically stored. This not only means that the operating system can move code and data in physical memory as it likes, but it also means that code and data don't need to be stored in physical memory at all!

A computer's processor can only access code and data that resides in physical memory (RAM). However, physical memory is relatively expensive so most computers have relatively little of it. Most multitasking operating systems extend their virtual memory management schemes to compensate for this scarcity of physical memory. They rely on a simple, but very important fact: Code and data only need to be in physical memory when the processor needs to access them! When not needed by the processor, code and data can be saved temporarily on a hard disk (or other device with abundant storage). This frees physical memory for use by other code and data that the processor needs to access. The process of temporarily transferring code and data to and from the hard disk to make room in physical memory is called swapping.

Swapping is performed to increase the amount of virtual memory available on the computer. The memory manager performs swapping "behind the scenes" to make it appear as though the computer has more physical memory than it actually does. Effectively, the virtual memory available on a computer is equal to its physical memory plus whatever hard disk space the virtual memory manager uses to temporarily store swapped code and data.

Loading Swapped Code And Data On Demand

If an application attempts to access code or data that is not in physical memory (it was swapped to disk) the virtual memory manager gets control. The virtual memory manager locates (or creates) an available block of physical memory, and copies the required code or data into the block so it can be accessed. Applications are not aware that their code and data were ever swapped to disk. The code and data are automatically loaded into physical memory by the virtual memory manager whenever the application needs to use them.

Key Points:

• Operating system memory management abstracts physical memory application and enforces isolation
• The memory manager extends the system's memory capacity via swapping

Ok, now that we’ve discussed how virtual memory and paging works, let’s relate this to virtualization.

Static Memory & Guest Only Paging

Today with Hyper-V (V1 & R2), memory is statically assigned to a virtual machine. Meaning you assign memory to a virtual machine and when that virtual machine is turned on, Hyper-V allocates and provides that memory to the virtual machine. That memory is held while the virtual machine is running or paused. When the virtual machine is saved or shut down, that memory is released. Below is a screenshot for assigning memory to a virtual machine today:

 

This memory is 100% backed by physical memory and is never paged. Remember that the guest is actively determining which pages should and shouldn’t be paged as it manages all the memory it’s been allocated by the virtual machine and it knows best how to do so. Here’s a basic picture to illustrate what this looks like in a virtualization environment. There are four virtual machines running and each of the guest kernels are managing their own memory.

Ok, now let’s dive into Second Level Paging…

Second Level Paging: What Is It?

Second Level Paging is a technique where the virtualization platform creates a second level of memory abstraction and swap files are created by the virtualization layer to page memory to disk when the system is oversubscribed. With SLP, you now have two tiers of paging to disk one within the guest and one below it at the virtualization layer. Here’s another picture to illustrate how Second Level Paging fits in. Again, there are four virtual machines running and each of the guest kernels are managing their own memory. However, notice that below them is the Second Level of Paging managed independently by the virtualization platform.

One common argument used in favor of Second Level Paging I’ve heard is this “If Windows and modern OSes all use paging today, why is this bad with virtualization?"

Great question.

Answer: Performance.

With Second Level Paging, memory assigned to a virtual machine can be backed by memory or by disk. The result is that Second Level Paging creates issues that are unique to a virtualized environment. When the system is oversubscribed, the virtualization layer can and will blindly and randomly swap out memory that the guest is holding even critical sections that the guest kernel is specifically holding in memory for performance reasons. Here’s what I mean.

Swapping the Guest Kernel

Swapping the guest kernel is an example where virtualization is creating an issue that doesn’t exist on physical systems. In an OS kernel, there are specific critical sections in memory that an operating system kernel never pages to disk for performance reasons. This is a subject where Microsoft and VMware agree and VMware states as much in their documentation.

“…hypervisor swapping is a guaranteed technique to reclaim a specific amount of memory within a specific amount of time. However, hypervisor swapping may severely penalize guest performance. This occurs when the hypervisor has no knowledge about which guest physical pages should be swapped out, and the swapping may cause unintended interactions with the native memory management policies in the guest operating system. For example, the guest operating system will never page out its kernel pages since those pages are critical to ensure guest kernel performance. The hypervisor, however, cannot identify those guest kernel pages, so it may swap them out. In addition, the guest operating system reclaims the clean buffer pages by dropping them. Again, since the hypervisor cannot identify the clean guest buffer pages, it will unnecessarily swap them out to the hypervisor swap device in order to reclaim the mapped host physical memory.

Understanding Memory Resource Management in VMware ESX Server p. 9-10; http://www.vmware.com/resources/techresources/10062

Thus, the more you oversubscribe memory, the worse the overall performance because the system has to fall back to using disk and ultimately trade memory performance for disk performance. Speaking of comparing memory to disk performance...

Memory vs Disk Performance

Finally, there is the performance comparison, or, really lack thereof, because there is no comparison between memory and disk. This isn’t debatable. This is fact. Let’s do a little math. Let’s assume that the typical disk seek time is ~8 milliseconds. For memory access, here are the response times in nanoseconds:

• DDR3-1600 = 5 nanoseconds
• DDR3-1333 = 6 ns
• DDR3-1066 = 7.5 ns
• DDR3-800 = 10 ns

So, if you want to compare disk access to DDR-3 1600 memory access the formula is .008/.000000005. Here are the results:

• DDR3-1600 memory is 1,600,000 times faster than disk
• DDR3-1333 memory is 1,333,333 times faster than disk
• DDR3-1066 memory is 1,066,666 times faster than disk
• DDR3-800 memory is 800,000 times faster than disk

We’ve heard on many occasions that virtualization users have been told that performance of Second Level Paging “isn’t that bad.” I don’t know how anyone can say with a straight face that a performance penalty of greater than six orders of magnitude isn’t that bad. To put 1.6 million times faster in perspective, assume it took you an hour to walk one mile. If you traveled 1.6 million times faster, you could roughly travel to Saturn and back in an hour. (Saturn is approximately 746 million miles away at its minimum distance to the Earth.)

Microsoft & VMware Agree: Avoid Oversubscription

The fact that swapping to disk carries a significant performance penalty and you should avoid it is another area where Microsoft and VMware agree. This isn’t new guidance so I’ve included examples from ESX 3 and VSphere.

From VMware:

Example 1: Make sure the host has more physical memory than the total amount of memory that will be used by ESX plus the sum of the working set sizes that will be used by all the virtual machines running at any one time.

--Performance Tuning Best Practices for ESX Server 3

Example 2: if the working set is so large that active pages are continuously being swapped in and out (that is, the swap I/O rate is high), then performance may degrade significantly. To avoid swapping in specific virtual machines please configure memory reservations for them (through the VI Client) at least equal in size to their active working sets. But be aware that configuring resource reservations can limit the number of virtual machines one can consolidate on a system.

--Performance Tuning Best Practices for ESX Server 3 page 15

Example 3: ESX also uses host-level swapping to forcibly reclaim memory from a virtual machine. Because this will swap out active pages, it can cause virtual machine performance to degrade significantly.

--Performance Tuning Best Practices for VSphere page 23

Translation: Ensure that the memory used by ESX and its virtual machines reside in physical memory and avoid swapping to disk, i.e. avoid oversubscribing memory.

Final Points on Second Level Paging

• Second Level Paging breaks the fundamental assumption that Guest Operating Systems have an accurate representation of physical memory
• Performance of memory to disk ranges from 800,000 times to 1,600,000 times faster
• When the system is oversubscribed, Second Level Paging carries a significant performance hit. Simply stated, the more the system is oversubscribed, the more it relies on swapping to disk and the worse the overall system performance.

The good news is that there are other ways to pool and allocate memory and Hyper-V Dynamic Memory is a good solution for desktop and server operating systems... In my next blog, we’ll explain Hyper-V Dynamic Memory.

Cheers,

Jeff Woolsey

Principal Group Program Manager

Windows Server, Virtualization