Remarks by Kirill Tatarinov, Corporate Vice President, Enterprise Management Division, Microsoft Corporation
Microsoft Management Summit 2005
Las Vegas, Nevada
April 19, 2005
ANNOUNCER: Ladies and gentlemen, please welcome Corporate Vice President for the Microsoft Corporation, Kirill Tatarinov. (Applause.)
KIRILL TATARINOV: Well, good morning, everyone. I’m very excited and honored to welcome you to the Microsoft Management Summit 2005. In the course of the last four years, MMS has emerged as one of the premier events in the systems management industry. In fact, it’s the only event that has grown approximately 50 percent in the last two years. And before I begin I want to thank your Gold Partners who helped us sponsor this event, our Gold Sponsors, Dell, Hewlett-Packard and SMS Alliance.
Now, some of you may have heard people refer to systems management as one of the most boring disciplines within enterprise IT, and I obviously have a different opinion having spent 15 years of my career working in systems management, and I know there are at least 2,600 people here in the room who would share that thought of mine and there are also many people who’ve tuned to this webcast today.
So why am I so excited and bullish about systems management? Well, the key reason is because systems management can truly help the IT, the enterprise IT to overcome the crisis of complexity and cost. And there is indeed a crisis of complexity and cost if you look at these numbers, if you look at how much the industry spends on managing of their assets in the enterprises. And this number is expected to grow and only a small fraction of this number is actually spent on software.
So there is a dramatic opportunity that we have here to change the landscape in the industry, to change the pace, to reduce the cost and to help enterprises take full advantage of innovation in the enterprise software.
One reason why people may think that systems management is boring is by simply where we spend our time. We all spend a significant portion of our time in maintaining existing systems. We all spend significant portions of our time in being very reactive and fighting fires, and how exciting is that?
So two years ago at this very venue, at the Microsoft Management Summit 2003, we at Microsoft set the course to change that, to change the landscape of systems management. Two years ago we announced Dynamic Systems Initiative, and two years ago we basically articulated the course of dramatic increase in the investment that we were planning to put across the Microsoft product line and the industry to make systems management something that will help reduce cost in enterprise IT and become an asset for enterprise IT.
So in the next 72 minutes I will update you on the progress that we made in the last 12 months delivering products in the systems management space, filling the roadmap of Dynamic Systems Initiative, I will give you an update on DSI and I will show you how DSI is becoming much more concrete with technology and the feedback that we’re getting from the industry.
I’ll show you our roadmap for the next two years. I know it’s very important for you to know where we’re going so you can make appropriate planning and investments so we’re going to be very open with our plans and with our directions for the next 24 months.
And finally I will share my thoughts on why I believe Microsoft is uniquely positioned to your partner in solving enterprise management needs of your enterprise.
So let’s start with the progress report. Let me talk to you about some unique products and technologies that we delivered to you in the last 12 months.
Microsoft Operations Manager
We love our MOM. This is a great technology, this is a great product. It’s been on the market since October of last year. There is great customer feedback that we’re getting, we’re very excited about the number of customers that are running Microsoft Operations Manager in production today, the number of new customers that signed up to run Microsoft Operations Manager since the product was shipped in October of last year and most importantly the market share of a product that came from behind, a product that came from essentially nowhere to now generates approximately 20 percent market share in the industry.
Microsoft Operations Manager is a very significant element of Dynamic Systems Initiative. Microsoft Operations Manager is what helps us and our industry partners today to take the knowledge about how our systems and our industry partner systems need to be run and operated, put that knowledge into models and deliver it to you in the form of MOM management packs, and that dramatically increases operational efficiencies through delivering this knowledge, through getting this knowledge.
MOM 2005 also has a continuous investment model where basically while the core technology of MOM was shipped last year we continue to innovate and industry partners continue to innovate by delivering those management packs to you on an ongoing basis.
In the last two weeks we announced eight additional management packs for MOM 2005. We made them available for download and we’re essentially announcing them at this conference. Some of those management packs target very specific technologies within the Windows and Microsoft product portfolio. Two of them I want to highlight because they really position Microsoft Operations Manager in the new light and really enable you to use this existing technology for two new scenarios.
First of all, we’ve extended Microsoft Operations Manager 2005 and enabled it to monitor mission critical desktops in your organization. The fact that MOM 2005 agents is very lightweight technology essentially enabled us to capitalize on that and give us the opportunity to deliver the management pack for XP to take advantage of the capabilities.
We also delivered a management pack for monitoring of Web sites and Web Services. This is essentially our implementation of synthetic transition technology, technology that continuously pings the Web site and gives you an idea of how the end users will experience the performance of that Web site.
Now, I realize many vendors in the industry would deliver these technologies to you for a significant premium beyond and above what you’re being charged for systems management. With MOM 2005 pricing and licensing model you just get these technologies with a base Microsoft Operations Manager license, which makes it the most cost effective solution in the industry.
I’m also very excited, very, very excited about the ecosystem of partners that came around to work with us on MOM. There are many of the announcements that are happening at this conference today; in fact, approximately 25 partners come together and make significant announcements on their investment to work with us to deliver complete solutions to all of you.
One of those announcements I want to highlight here. This announcement comes from a relatively small company called AVIcode. And what they’re delivering and they will be delivering later this year is a solution built on top of Microsoft Operations Manager 2005 to help you monitor custom .NET applications. This solution will come in the form of a management pack for MOM and then AVIcode will deliver extensive solutions to build on top of that to allow you to diagnose your applications.
So very exciting deliverables in the MOM area, very great things that are happening and continuous investment that will help you get even more value out of MOM 2005.
Windows Update. We’ve been talking to you about patch management, we’ve been talking to you about the work that we continue to do to help automate patching, to help make the patch nightmare easier. Windows Update is a prime technology that we use in that area and investments in Windows Update are what helps us keep your machines up to date.
There are some facts that I wanted to share with you today, some facts that are not widely known in the industry and I think it’s important to understand the scale, the scale of Windows Update and the scale of the help, the helping hands that we deliver with Windows Update technology.
Two hundred million machines are kept up to date with Windows Update every day. This is a phenomenal number. I don’t think there is any other site that connects to 200 million machines and sends updates to those machines.
And it’s not just small updates, it’s also updates of large scale such as XP, Windows XP Service Pack 2; 120 million machines were successfully updated to Windows XP SP 2, which is the most secure desktop operating system, with Windows Update and most of those machines were updated in an automated fashion, which means that very little time was spent by the end users and by your users to execute this update.
Windows Update also comes in a federated server fashion that can be deployed by small and medium organizations to help them gain more control on when patches are delivered. This technology is called Software Update and we’re very excited that approximately 150,000 — 150,000 small and medium organizations have successfully deployed and used this technology.
We’re readying updates for Windows Update technology and for Software Update Services. Windows Server Update Services is currently in beta and it brings even more simplicity and more ease of use and more control to existing Software Update Services. We’re very excited about this beta, it’s in its final stages, the product is rock solid and I encourage you to take a look at it, it’s available for download.
And another very significant investment in the area of Windows Update is the fact that Windows Update in a short two months will become Microsoft Update and essentially enable us, in addition to all the critical updates for Windows, to post updates to other technologies from Microsoft such as Office, Exchange, SQL, Visual Studio and many others, basically giving you an automated way to receive updates for all technologies from Microsoft.
So very exciting things happening in Windows Update and patch management is very critical for us. (Applause.) Thank you.
Systems Management Server
My job has gotten much easier in the last year. and I know many of you would share that thought; well, yours for different reasons because of the technology, mine is because I don’t have to go around the world and apologize for mistakes that we had with SMS version 1 and version 2. (Laughter, applause.)
SMS 2003 is rock solid. It’s been on the market for 18 months now, Service Pack 1 has shipped, a couple feature packs have shipped, we know that this is a super solid foundation for you to run changing configuration management in your enterprises. We’re very proud of this product.
We’re also very proud that the industry has recognized what we’ve done to improve SMS. There is a significant number of awards and accolades that SMS received from the industry. You should be proud that you’re running the champion in your enterprises.
There is a growing list of customers of SMS. Over 20,000 customers, unique enterprises run SMS in their environment to basically run their mission critical enterprises.
And the other significant figure I wanted to recognize, 83,000 unique downloads of Service Pack 1 that was made available in the last fall, which means that people don’t just buy SMS but actively use it and actively download new versions and new feature packs for this technology.
Obviously we have something new coming for SMS as well. With the release of Microsoft Update and Windows Server Update Services, we now have a unique opportunity to fully integrate scanning and updating technologies that are available from Microsoft. This is something that we committed we would do for the industry and this is something that we’re starting to deliver on with an update to SMS 2003 to be available shortly after Windows Server Update Services will ship.
We also followed your feedback and we’re readying minor updates to SMS 2003 to be available in approximately 12 months time and in addition to various small things that we will add to the products there are two very important scenarios that we’re making available. First it’s vulnerability assessment: Basically taking advantage of existing SMS inventory we’re now giving you the ability to assess the vulnerability and how your enterprise can potentially be penetrated by intruders.
We’re also extending the capabilities that we built with our partners, specifically with Dell, to open up SMS to enable you to bring patches and updates from a variety of industry partners. So it’s not just Microsoft download center that SMS talks to, not just the Microsoft Windows Update site, it’s also our partners, various ISV and hardware vendors who could use, leverage SMS technology to bring their updates to the enterprise and help you implement it in the right fashion.
Microsoft System Center
Microsoft System Center. We announced Microsoft System Center approximately two years ago and we’ve been busy working to evolve the roadmap to understand exactly what you need us to deliver in that area. We made our thinking much more concrete, we now have a firm roadmap and we now all collectively understand what System Center actually is.
System Center is a family of management products from Microsoft. The feedback that we received from you is that you would like to get technology that is integrated but yet modular, enabling you to preserve your investment in existing products such as MOM and SMS but yet add value beyond and above it to help you leverage data that already exists in MOM and SMS and gain more capabilities, gain more functionality.
And that’s exactly what we’re planning to do with Microsoft System Center. When we talk about the roadmap you will see specific technologies delivered and integrated in yet very modular fashion that will give you an opportunity to take full advantage of your existing investment and yet gain more capabilities from existing products coming in that domain, in that environment.
Another very important point of feedback that we’ve been receiving from you is in the area of heterogeneity. It’s quite obvious that there is no such thing as a homogeneous enterprise. As much as we wish there was one, there isn’t. There are other platforms, there’s a lot of legacy, there are a lot of systems coming from other vendors; we fully recognize it.
And we’ve made significant investment in the area of embracing heterogeneous environments. We did it in three areas. First of all, we made our own servers as scalable and as interoperable as we could to make sure that we can actually bring data from non-Windows platforms and fully accommodate it in our server technologies.
Second, we worked with management partners. We made significant investment in growing partner ecosystems and we made a significant investment in interoperability to make sure that we can bring data from non-Windows environments, either coming to Microsoft products from other management technologies or coming from unique partners like Vintela who helps us extend SMS and MOM to directly bring data from non-Windows environments into MOM and SMS.
We also made very, very exciting investments in the standards, in the area of standards. I mean, truly let’s recognize, let’s face it, if we lived in the ideal world where every system knew how to interoperate with each other, where every system, whether it’s management system or managed system would speak the same language we wouldn’t have a problem of heterogeneity, everything would be speaking the same language. Well, that’s exactly the aggressive goal that we set for the industry and for ourselves, to achieve this ideal world where everybody is speaking the same language.
We obviously fully recognize that this undertaking is not something a single vendor could solve by himself, even a vendor like Microsoft. We partnered very deeply with Intel and we coauthored WS-Management, which is the new standard, the new protocol, the new transport protocol for communicating of systems management data across Web services. We were later joined by Dell, Sun and AMD, and we together published the specification, which is available and which is being implemented by a variety of industry players.
So we’ve essentially made one step toward the goal of getting to the world where everything is speaking the same language. Now, I’ll be the first to admit that it’s going to take a few years before this technology will be fully implemented and before we get complete interoperability as it relates to systems management, and while we do it we’ll continue to invest in our partner ecosystem and we’ll continue to make sure that you can take full advantage of enterprise management technologies from Microsoft and get it extended to the enterprise.
I would like to specifically call out some partners that have worked with us, the closest and some very unique announcements that they made in the last six months in the area of working with Microsoft and embracing WS-Management and other management technologies from us.
First of all is Intel. As I mentioned, Intel essentially coauthored, co-developed the WS-Management spec with us. I was super excited about the level of engagement that we had with Intel and the amount of contribution by them. Intel has also committed to implement WS-Management in their active management technology, which is essentially a chipset allowing to better manage Intel-based motherboards and hardware. This is very, very important to understand that no standard in the industry can be successful until it’s implemented. And it’s not just that we delivered and published the standard, we’re also getting commitment from the industry to implement this.
Dell, another coauthor in WS-Management, they’re working on their commitment to implement WS-Management in their server line and there’s also very, very unique work that Dell has done with us in the last six months in the area of SMS and MOM. Dell partnered with us to deliver an inventory tool for SMS that enables you to scan Dell servers and give you an idea of what patches you may need to download from the Dell download center. This technology was made available in January of this year and in the last three months we had almost 25,000 downloads, which means that there’s huge interest in this area, huge interest from our customers, from all of you to see hardware patches and software patches coming together through one technology and that’s what this technology enables.
Dell has also developed a management pack for MOM to monitor unique characteristics of their servers and most excitingly Dell has OEMed MOM Workgroup Edition. And this week Dell is actually starting to take orders on their servers shipped with MOM Workgroup Edition. So next time you call Dell to purchase a server ask for MOM Workgroup, it’s going to be there.
Hewlett-Packard. We’ve made significant progress working with Hewlett-Packard in the last year. Just like Dell, Hewlett-Packard is making commitments to deliver inventory tools for SMS to scan for specific updates for Hewlett-Packard firmware and BIOS. Hewlett-Packard also delivered a management pack for MOM and I would say that they have done a very good job in building deep knowledge of their hardware into models and delivering it through MOM management pack. I’m very excited about the work that HP has done there and I’m very thankful for them helping us sponsor this conference.
And last but not least is IBM. We’ve come to agreements with IBM on many very strategic topics; we’re working on many others. IBM is working to integrate IBM Update Express with SMS and IBM is also working to integrate IBM Director with MOM to yet again extend MOM with specific hardware data coming from IBM servers.
So it’s very exciting work in partner areas, very excited about the progress and you will see through the rest of my talk on how those technologies actually help to complete the management scenario and to deliver it end-to-end.
So great progress, great progress in the last year and I hope you appreciate how busy we’ve stayed to deliver on what we committed to you.
Running an Enterprise-Class Network
MOM 2005, I wanted to say a couple more words. When we first announced the MOM 2005 plan back two years ago at this conference, at this very room, we set a very aggressive goal to make MOM an enterprise class management product, not just a little Windows manager, but enterprise-class product, a product that can manage enterprise end-to-end, a product that can take events and alerts from any device into the network; a very aggressive goal.
And today I’m very proud to show you how that goal was achieved. What you see here on the screen is the diagram of the network that runs this event. There are a variety of networking devices, there is a variety of wireless devices, all the labs, comnet, demos, all managed by this network. And what I would like to show you is how MOM 2005 runs this network today, runs the network operation console from top to bottom, taking advantage of deep partnerships we’ve built with companies like JalaSOFT and Hewlett-Packard and provide end-to-end management of this event. I would like to invite Vlad on stage who will help us with this demo. Vlad. (Applause.)
VLAD: Thank you, Kirill. Well, Kirill, I’d first like to start by showing you the state of the servers that we’re managing right here in comnet. I’m going to do this by showing you our MOM operator console. Here on the left you see all the different views that I can choose to select, these views down here came with the management pack that we have imported. Along the top you can see that I have the console scoped to a group called the Windows Servers Group. We’re going to change other groups later but right now I’ll leave it in the Windows Servers. And along the right you see all the tasks that I can use to resolve problems when the state of my servers change.
I’m going to take away the task pane to maximize this main window and here you can see the state of the servers that I’m managing along the left and along the top here all the roles that these servers are playing. Now, these roles are discovered by the management packs we’ve got imported. We’ve got the Active Directory role, the HTTP role. If I scroll over to the right you see the ISA role, the MOM agent role, Operating System role, the SMS Server role.
Now, when I select a particular role on a server like this Exchange role, I can see the components that make up the Exchange role on this server and the state of these components, as the management pack is defined. I can select the disk role, looks like all the disks on this particular machine are doing okay. I can select a particular AD role. Now here I can see all the components that make up Active Directory. These are again things that the management packs are doing to make sure that the replication health is successful, all the services that make up Active Directory are successful.
I’m going to switch and show you the topology view for Active Directory so I can sort of see what the context of my Active Directory domain looks like. I expand the Active Directory view on the left and now I expand the replication topology diagram and I’m going to select the connection objects. Here you immediately see the two servers that make up our Active Directory domain. You can see that they’re both global catalogue servers by the phone book icon. Now, I can hover over a server to see more details.
It’s definitely something that we’re going to have to investigate after I finish this keynote, I’ll have to go back to the comnet and see if it’s impacting our users.
I can also hover over a particular connection object. Right now I know these connections are successful because they’re green but I can see all sorts of details about this connection like what the last successful synch time was, what partitions are held by this connection object.
Now, as you earlier mentioned, Kirill, we just released eight new MOM management packs that we have imported and we’re using here in comnet. I’d like to show you two of them. The first is the Web Sites and Services Management Pack that I’ve got configured here. I’m going to expand the Web Sites and Services view over here, expand this particular group and I’m going to click on the response times.
Now, I’ve configured the Web Sites and Services Management Pack to manage our internal comnet Web site. This is that Web site that you see automatically logged in when you log into any comnet machine. Now, I’m testing the response time of that Web site and I’m going to view this graphically and we can see what the last value was. And if I view this graphically by clicking Draw Graph here I’ll see the performance of this Web site over the last two hour period. And it looks like pretty much we’re having response times between 12 and about 40 milliseconds so things are pretty good. We’ll keep track of this as lunch happens or between sessions when people are logging in to comnet to see if there are any weird numbers here or any increases in response times that we may want to look at.
The next I’m going to show you is the Desktop Management Pack. I’m going to expand the Desktop view on the left and I’m going to click the state view. Now, right now you don’t see any servers here or any desktops here because I’m scoped to the servers group. If I drop this box down and select the Windows desktops, I can immediately see all the desktops that we’re managing here in comnet. These are again the desktops that you’re logging into on the comnet site. I can see all the roles that make up these different desktops, I can look at the disks, I can look at the MOM agent roles, the operating system roles. We’re even monitoring the SMS client that’s installed on those different desktops.
Now, I know that you NOCs are a lot like our NOCs here in that you don’t just have Microsoft operating systems and applications; you have hardware that you need to monitor, you have devices that you need to monitor and you have other applications that you need to monitor. I’d like to highlight two of our partners that we’re using in comnet. The first is the HP ProLiant Management Pack. If I expand the HP ProLiant servers view on the left, click the state view, I need to switch back into the Windows Servers Group and now I can see all the servers that we’re monitoring using the HP ProLiant Management Pack. I can see the roles that the HP Management Pack has defined, I can see the health of the components that the HP Management Pack has defined. These aren’t things that I had to customize and configure; I simply imported the management pack, MOM figured out the rest with the management pack configuration information.
The next thing I’d like to show is how we’re managing all the different network devices in this particular environment. Now, this includes the network devices in comnet that allow you to get out to the Internet, includes all the network devices that provide the network to all the different session rooms; all of this we’re managing using the Xian Network Manager provided by JalaSOFT.
To show this I’m going to change our group now to a group called House Network. This is a group that we’ve defined to include all the different devices that we’re managing and it definitely looks like there’s a number of different devices that we’re going to have to investigate since there’s a bunch of errors and warnings that someone is going to need to review to make sure they’re not critical and you can see we’re managing APC UPS devices, we’re managing Cisco switches and a number of different network devices.
I can select on a number of these different devices, see the components that make up that particular role. I can even click on the diagram view and view kind of the hierarchy view that we’ve created on this particular device group.
Now, Kirill, for fun I’d like you to unplug one of these network cables.
KIRILL TATARINOV: What is this box, Vlad? Is this one of those devices?
VLAD: This is one of these devices. One of these devices, it’s actually this one right here and I’d like you to unplug a particular network cable. This is a device that’s being managed right now by the Xian Network Manager that’s integrated into MOM, using their management packs and their integration using the other MOM APIs.
Now, in just a few moments we’ll see this state updated and we’ll see the overall state change to red and then it will flow up to the overall group and you can see that now this server is red, the group that it’s in is the MMS comnet, so this device we’ve grouped into the MMS comnet group and it’s flowing and making the overall house network red.
I can right-click on this device, click view, show alerts, I can see that alert that’s available on that device, I can view the data about the alert, we can resolve it, we can assign it to a user, we can do all those sort of things. But the really interesting thing now is now network device information and application information is in MOM, we can start to put service groups together that really provide our overall state or our service, which is a combination of both the application and the different devices.
Now, Kirill, we have barely scratched the surface of all the different ways we are using MOM in comnet. If you’re interested in learning more, I encourage you to stop by the NOC, which is in the comnet room, check out the four plasma screens that we have displaying active views showing you the state and reports of our partners and our different management packs. And we also have a session on Thursday where we’re going to spend an entire hour talking about how MOM is managing comnet.
Thank you very much.
KIRILL TATARINOV: Very good. Great demo. Thank you, Vlad. (Applause.)
So I’m very excited about our ability to reach this aggressive goal of becoming true enterprise management in a true sensible world so we can now rival the incumbents in the enterprise management domain. It’s also very exciting to see how the experts who have the knowledge about individual domains put that knowledge into management packs and deliver them to you and in this particular demo JalaSOFT was the expert in the networking domain and Hewlett-Packard was the expert in the server hardware domain.
Dynamic Systems Initiative
Now let me switch gears and give you an update on Dynamic Systems Initiative, our vision and our strategy on a lot of things coming together.
When we talk about Dynamic Systems Initiative we need to start with the lifecycle, the lifecycle of potentially every system. If you think about distributed computing and the amount of complexity that exists there today with a large amount of different moving parts, one thought comes to mind and one thought really drags down systems management historically and the thought is you cannot manage the unmanageable. A system that is not designed to be managed in the long run cannot be properly managed. You can attempt to reverse engineer, you can bring technology on top of it to try to bring the manageability later on but fundamentally it’s not going to work, fundamentally it’s just going to bring more complexity.
So the key premise here is we have to bring the links to this lifecycle, we have to connect the developers who build the systems to the people who runs the systems so there is knowledge that exists and there is knowledge shared.
The lifecycle, the complete lifecycle actually also includes the end user, the end user who gets the full benefit of the system as it gets deployed, and that end user needs to be able to interact, needs to be able to communicate with both the IT professional who runs the system and the developer, proactively communicating their feedback, proactively and programmatically letting them know whether the system performs what it’s supposed to do or whether it doesn’t. And that lifecycle, which is for the most part broken today, that lifecycle needs to be connected, that lifecycle needs to come alive, truly alive for the IT professional, for all of us to regain control of the enterprise, for us to become more proactive, delivering value to our end users.
Now, we talk about knowledge in the context of Dynamic Systems Initiative and knowledge does indeed exist in every step of the way. Most developers have prescriptive guides, best practices and implementation guides basically explaining how the systems need to be built.
Every IT professional, every operational side of the house that I’ve visited in the last few months, has a certain amount of knowledge and best practices and policies and guides and methodologies developed. Well, today they either exist in the head of the IT professionals or they exist in those little books, written words on how the system needs to run.
And the end users typically have definitions of Service Level Agreements, Service Level Agreements that specify what kind of level of service they expect from IT professionals and what kind of applications they specify they want to be delivered from application developers.
But those sets of knowledge today exist in isolation, those sets of knowledge are not connected and not linked and they’re not programmatic. And the main job of DSI, the main job of DSI is to bring that knowledge together, to express this knowledge in models and to make those models available across the application lifecycle. So it is knowledge expressed in models across the application lifecycle, that’s what DSI is all about.
And the goal that we’re accomplishing with that and the main driver for what we do with DSI is cost reduction and increased productivity. We’re investing in DSI and we’re investing in systems management to increase operational efficiency, to optimize your infrastructure, to give you a chance to take full advantage of the investments in enterprise IT, whether that enterprise IT came from Microsoft or the rest of the industry.
And there are obvious benefits to this connected lifecycle and knowledge expressed in models that come to every persona across this lifecycle. Application developers can significantly reduce time to building and deploying the applications that they’re building. The IT professional, people on the operational side of the house finally get back in control, knowing what’s coming to them from the application developer side of the house and from their vendors and also being able to proactively respond to the end user feedback that comes in a programmatic fashion. And, of course, the end users get one step closer to the dream of getting computing as a utility, getting computing in the form of a dial tone that just works for them and does what it’s supposed to do.
So that in essence is DSI and that is the core of the DSI strategy, knowledge expressed in models across the application lifecycle.
System Definition Model
The core technology foundations for Dynamic Systems Initiative, the most important technology component here is System Definition Model. This is, in fact, that common modeling language that we use and we share with the industry that enables us and the industry to express the knowledge into models, to put the knowledge in the model.
Technically, System Definition Model is essentially common schema, it’s an XML document that enable you to put definition of the model into the system and make our products know what the model is out there.
The second very important technology component and the one we talked about already is WS-Management. Basically think about System Definition Model as the payload and WS-Management as the transport. System Definition Model defines what is the managed environment and how it needs to be managed and WS-Management enables you to connect that managed environment to management systems and also connect multiple management systems.
There is also a very important aspect here, the aspect of federation and the aspect of replication. There has been a lot of talk in the industry about change-management databases and the fact that there’s got to be one single change-management database that sits somewhere in the core of the enterprise.
While we certainly agree the change-management database is a very important aspect of the enterprise IT, we don’t quite believe in the central, one single point of failure notion of CMDB that’s being presented to the industry. We believe in dynamic CMDB and we believe that by making available these common schema and by making available the standard communication protocols we’re going to be able to implement dynamic CMDBs that on one hand will provide complete connectedness of all the managed and management environments and at the same time avoid the single point of failure and put the data locally to where the data needs to be.
So that’s the technology foundation of DSI, System Definition Model, WS-Management, two core building blocks of what we’re doing.
Microsoft Operations Framework
Now, technology is very important and essential but not sufficient for something as complex as systems management to be fully implemented and deployed in the enterprise. And following that thought we’ve been making significant investments in our methodology and we continue to make investments in the Microsoft Operations Framework. The Microsoft Operations Framework is essentially our implementation of ITIL with extensions that we brought to the industry.
ITIL is one of the best methodologies available in IT today to help you run your IT and essentially identify how enterprise computing environments need to be measured from the process standpoint.
We adopted ITIL, we expanded ITIL, we added models that you see in front of you to ITIL to basically identify core disciplines — change management, operations management, supporting and servicing and customization.
And we also made ITIL much more prescriptive by adding prescriptive guides and best practices, sharing experiences from the other users and from ourselves with our customers.
The Microsoft Operations Framework is available today, it’s available for download. We link it very tightly with our products and, in fact, it goes very deep into our product developers who build products with Microsoft Operations Framework in mind.
And we also work very closely with the ITIL community, basically working to contribute everything that we have done to extend ITIL back into the community and help the community create new versions of ITIL with the extensions that we’ve built. And that methodology essentially is going to help IT professionals to fully implement the Dynamic Systems Initiative and take full advantage of it in a much more rapid, prescriptive and guaranteed fashion.
Now, the roadmap. This is probably to me the most exciting part of this presentation to tell you what we’re going to be giving you in the next 24 months. We showed you a similar set of slides, rewind back 24 months in the conference in 2003 and we’ve delivered on everything we told you and today we’re sharing you what’s coming to you in 24 months ahead.
First, a very important addition to the Microsoft management family and the first product in the System Center product family is System Center Reporting Manager. This technology essentially integrates data from MOM, SMS and Active Directory. This technology has extensible schema enabling you to bring your own data into the data warehouse and this technology also brings a very well defined set of reports, which enables you to make better business decisions on top of those datasets.
Of course, System Center Reporting Manager brings the most value if you have SMS, MOM and Active Directory but not necessarily. Even if all you have from our management product portfolio is SMS you can still gain value from System Center Reporting Manager by essentially consolidating multi site view and putting it into offline data warehouse. And if all you have is Microsoft Operations Manager you can also still take advantage of System Center Reporting Manager by integrating data across the multiple site hierarchy.
We’re super excited about this product. It’s going into beta in just a short two weeks, so early May you will have a chance to download this technology and start using and start playing with that; very excited about the advances and innovation in this area.
The second System Center product on our roadmap is System Center Data Protection Manager. I’m very excited about this product for several reasons. First of all, it’s a significant step in extending our portfolio of products. Basically with System Center Data Protection Manager we’re entering the market of backup and restore and it signifies our ambition to go broader from traditional MOM and SMS areas into additional disciplines in systems management.
I’m also excited about the way this product is built and the way it simplifies complex operations of backup and restore.
System Center Data Protection Manager is disk-based backup solutions. It’s also built on Microsoft technology of shadow copy, which means that the operation here is continuous, which means that you do not have this treacherous backup window where everything is stopped and you need to do backup and it also means the backup can be done very, very fast.
The fact that the product is built on a disk based technology also allows for very rapid recovery, high reliability and also delivers a very unique key scenario where administrators with the right privileges can actually do self recovery of files without having to go to the backup administrator at all, and that is dramatic reduction of costs.
The product is very easy to set up, the product, of course, comes with more management packs so it comes with centralized management and the product uses standard server hardware and software from Microsoft, it’s built on Windows Server 2003.
The product is currently in beta, the beta is available for download. If you didn’t have a chance to look at it I would encourage you to do it. And we look forward to delivering this product to you in the second half of this calendar year.
This technology is one that makes me really proud of working at Microsoft. I demoed this technology to you last year at the end of my presentation. This technology was called (Indy ?). Indy was the codename of this product. This technology originated in Microsoft Research, it made its way into the product group, approximately 18 months ago we made the decision to productize this technology.
And today I’m very proud to announce to you that this technology is coming to market this calendar year in the form of System Center Capacity Manager 2006. As we discussed last year, the first scenario that we’re going to be addressing with this technology is sizing solutions for Exchange and also Microsoft Operations Manager. It’s essentially going to enable you to do predictive planning of what Exchange deployments ought to be, of how you need to make architectural decisions to fully optimize Exchange or MOM deployment.
This technology models performance, it helps you identify future bottlenecks, essentially predict future bottlenecks before you go into user deployment and this technology will also enable you to predict user experience, basically response time that the end users will be getting from Exchange and MOM.
We’re very excited about it. Today, in addition to announcing the name and the delivery schedule for this technology, we’re also announcing a Community Technology Preview for our partners, which means that the partners will walk away from this show with the CDs with System Center Capacity Manager 2006, formerly known as Indy, in their hands, fully running, fully operational.
With that, I would like to invite Sabrena McBride on stage and Sabrina will show us System Center Capacity Manager 2006 in action. Sabrena, please. (Applause.)
SABRENA MCBRIDE: Thank you, Kirill.
You know, I’m really excited about demonstrating this very important edition to the System Center family. How many of you ever thought capacity planning would ever be easy? I think I see one hand way in the back over there. That is exactly what this product enables you to do: we make capacity planning easy for you. We basically, as Kirill stated earlier, will predict how an application will respond with the ever-changing demands in your environment, we’ll actually tell you in advance exactly what your potential performance bottlenecks will be, in addition to enabling you to try what-if types of scenarios prior to touching anything in a live environment in your test labs.
One of the most difficult tasks associated with capacity planning, as we all know, is actually capturing accurate data regarding the configuration information in our environment and utilizing that information to build a useful model.
What you’re seeing here is what we’ve done for you. We’ve actually captured that information, captured that knowledge, built a model of an Exchange environment and that Exchange environment pretty much consists of two regional offices, one in Seattle, one in New York, and they’re connected via a corporate WAN. We also have three branch offices — Redmond, Bellevue and Tacoma — that are connected directly to Seattle via point-to-point.
As you can see in the product, we’ll actually tell you how those users are accessing Exchange. So some of them are actually exclusively going through Outlook Web Access and we have others that are going through Outlook cache mode and, of course, the regions are connecting via connected mode.
So it’s really cool, right? If you don’t think that’s exciting enough —
KIRILL TATARINOV: This is an imaginary company that we model here and they take the model, they took the model from the Microsoft model that gives all the data about Exchange 2003 and built their own environment using this modeling.
SABRENA MCBRIDE: That is correct.
KIRILL TATARINOV: And they basically captured it into the product and now they look at the model as opposed to the physical environment as they plan the future.
SABRENA MCBRIDE: That is correct. Thank you, Kirill.
So if that’s not exciting enough to you, let’s just assume that you guys are the Exchange administrators and your company has now acquired a company out of Boston that has no messaging infrastructure whatsoever. So you are now tasked with quickly getting those users up and running on their e-mail, so you have to make a quick decision: do I build an infrastructure in the Boston office, knowing I don’t have expertise there, or do I go ahead and integrate them into my existing environment.
And what this product is actually going to allow you to do is simulate that process to determine what’s actually going to happen in your environment and are your end users going to experience the desired results.
One other thing that I just want to quickly show you, in addition to the topology mapping, we do give you all of the information with regards to your existing model that comes from your live environment. We’re actually going to go and edit that model to add a Boston office and the additional 1,000 users that are in that branch office.
So to do that we’ll first add the Boston office to our model and this just shows you how easy it is to change your model. You don’t have to be a model expert to do this.
KIRILL TATARINOV: You don’t have to be a developer either.
SABRENA MCBRIDE: Exactly.
We determined that since Boston is so close to New York, of course, it makes a lot more sense to connect them directly out of New York.
KIRILL TATARINOV: So we’re basically saying that this acquisition needs to come to the fold much quicker than time would allow us to deploy the new infrastructure in New York and we bring it through existing New York servers.
SABRENA MCBRIDE: Exactly.
KIRILL TATARINOV: Through Outlook Web Access.
SABRENA MCBRIDE: Exactly, Outlook Web Access and we’re actually going to be hosting their mailboxes on the New York server.
We do understand that there are a couple of risks involved in this. We don’t necessarily know that the New York servers can handle the increase capacity of a thousand new users, but in addition to that we want to make sure that those Boston users have a positive experience.
I’m going to go ahead and describe my Boston users within the model as well and, as Kirill stated, they are Outlook Web Access, through a fast Ethernet.
I’m going to quickly go back to my topology viewer just to make sure that the model is actually capturing exactly what I proposed for the environment, which it is.
KIRILL TATARINOV: So Boston is here.
SABRENA MCBRIDE: At this point I’m going to run a simulation. Now, what this is doing is it’s actually running tens of thousands of transactions against my current modeling environment that actually came from live data and the only modification that I’ve made is the addition of the Boston office and the thousand users.
Now, this summary actually gives us an idea of our overall system utilization and you quickly notice that you have a red indicator icon here that tells you that you have a potential problem on your disks in the New York office.
KIRILL TATARINOV: So those are the thousand extra users placed on the mailboxes obviously caused the extensive use of disk space.
SABRENA MCBRIDE: Exactly. And we don’t necessarily know what’s actually happening on those disks or what the purpose of that box is, so we can actually drill down a little further and take a look at — and I really just want to focus on New York for now and the disks, which is where I had my potential problems. So you can see the top utilized disks are directly associated with your mailbox store.
At this point there are a couple of decisions that you can make. You can decide, you know what, I need to add an additional server. That can be very complex, as you all know. Or you can make a quick and easy decision and say, you know what, maybe we should just add additional disk to that server.
Well, let me go back to my model and try that out and see how it affects my results.
So now I go back to my model editor and I’m going to edit server configurations within my model. I have my New York servers, the AD server, the Exchange server, and I actually have standard machines that I can implement Exchange on. I notice here that I have the exact same machine that I’m currently running Exchange on but it has eight disks instead of four. So we’re going to select that —
KIRILL TATARINOV: So there are standard configurations defined for my environment.
SABRENA MCBRIDE: That is defined automatically for your environment, exactly.
At this point I’m going to run the simulation again and I’m running the exact same workload against the exact same environment. The only change that I’ve made at this point is the hardware configuration.
KIRILL TATARINOV: So what would happen if there was a different server placed there. Okay, we fixed it.
SABRENA MCBRIDE: As we expected, as we hoped.
One of the things that we have not seen though, now that we have our infrastructure situation handled and we know that our current users are happy, the Boston users didn’t affect my infrastructure because of the change that I made, I haven’t taken a look at what is the actual experience that those users will have coming out of Boston. My goal is roundtable transactions will take less than one second to complete.
I notice here in my summary screen I do have some transactions that actually take longer than one second. Are those my Boston users? Well, let me drill down a little further and find out.
So this screen actually allows you to focus on transactions per site and at this point I’m really interested in the user experience for my Boston office. And as you can see, all of those transactions that are running out of the Boston office are completing in less than one second, which was basically our desired result.
KIRILL TATARINOV: So the goal is met and the architecture will work.
SABRENA MCBRIDE: The goal has been met, absolutely.
KIRILL TATARINOV: Very good.
SABRENA MCBRIDE: So as you can see, in less than seven minutes we’ve actually demonstrated how you can model your environment using System Center Capacity Management, how you can indicate what exactly might occur if you add additional users into your environment and how do you prevent that from happening and how can you validate any changes that you made before actually going into your live environment and making those changes.
KIRILL TATARINOV: Very good. Thank you, Sabrena, impressive.
SABRENA MCBRIDE: Thank you, Kirill. (Applause.)
KIRILL TATARINOV: So it’s a great example on how the knowledge about Exchange System, built by the Exchange team and Microsoft, is put and modeled, developed with the System Center Capacity Manager and delivered to our customers who they can build models for their specific environment, putting knowledge about that environment into models and put it all into action, basically another great example of DSI becoming live and delivering value today.
Microsoft Operations Manager Update
Now let me shift to probably the most exciting part of this roadmap discussion. Here I’m unveiling new versions of MOM and SMS. And I know most of you here are MOM and SMS users, so I’m very excited to talk to you about it.
Microsoft Operations Manager is codenamed version 3, basically the third release of MOM. There are three very exciting areas of investment that we’re putting into MOM. First of all, MOM will move to the concept of model-based management and through embedding models in MOM version 3 we are going to be able to deliver on a service-oriented monitoring set of scenarios, basically enabling you to define models of services that you deliver to your end users — and you saw the example in System Center Reporting Manager on how easy it is to define this model, you don’t have to be a developer to do it, and then monitor that service exactly the way it’s delivered to end users, not just individual servers or components but the service top to bottom. So it’s a very, very important area of investment, very important feedback that you’ve been giving us in the last year as we were introducing MOM 2005 to market. You truly wanted us to deliver it and we’re delivering it in the next version.
We’re also making investments in making MOM more extensible and also helping you gain more control from Microsoft Operations Manager. So we’re dramatically improving tasks and command control inside MOM.
We’re making investments in something that internally in Microsoft we call a “do-it” button. And basically when you see the knowledge that’s delivered with the management pack and when you see the recommendation coming from this knowledge, if those conditions happen and you see those conditions happen, this is the step that we recommend.
How many times did you wish to have this little do-it button that you could click and say, “Do it now?” Well that button is coming with MOM version 3 and it will help you be much productive and much more efficient.
We’re also working on much deeper integration with the platform. MOM version 3 will come to market in the approximate timeframe of the Longhorn release of Windows and we’re taking full advantage of investment in management technologies that go into Windows.
We are making investments in making the interface and access to MOM much more roles-oriented in that we’re building much closer linkage with Active Directory to basically help us define policies of individual MOM administrators and also linking MOM much closer with roles within Windows Server.
And finally, we’re also taking on probabilistic analysis in MOM 2005 where you’re essentially going to be able to predict cause analysis of potential future failures.
So we’re very excited about the innovation that’s happening in MOM version 3. We’re using this conference to capture feedback from some of you. We’re running some of the workgroups to capture this feedback to make sure that we address your needs with this next version of MOM.
Systems Management Server Update
SMS, very exciting innovations happening here. Software Management Server version 4, SMS 2003 was essentially version 3, this is the next release. Yet again we’re taking a model-based management approach and we’re embedding it in the next version of SMS.
And it will enable us to do desired configuration management. Basically through using models you will be able to express desired configuration of your environment and then using SMS version 4 you will be able to keep your environment in that desired state. And whether it’s your internal policy that you want to apply or some industry compliance guideline such as Sarbanes-Oxley or HIPAA or some other, SMS version 4 will be the tool for you to do it.
We continue to invest in security scenarios in SMS. With SMS version 4 we’re addressing Internet facing scenarios and we’re making significant investments to making sure that we do it in a very secure fashion and it’s a very important investment.
And we’re also making significant investments in integrating SMS version 4 with network access protection technology, also know as quarantine, which will be available in Windows “Longhorn,” and yet again it will continue to help you button down your environment, making sure that it’s more secure.
Last but not least, a set of features and a set of investments that I know many of you will be very, very excited about when you see that. That’s an investment in dramatic simplification of Software Management Server user interface, making it much more task-oriented, making it much more role-based, something that you asked us to do for a long time and something that we’re investing in and delivering with SMS version 4.
And with that simplification we’re integrating various technologies for deployment of operating systems, technologies specifically in the area of automated deployment services that we have available today, investments that we’re making in Windows Longhorn for imaging of the system, investments that we made with SMS OS deployment feature pack. All those investments will come together in integrated OS deployment delivered with SMS version 4.
So a very exciting set of innovations, yet again we’re running some focus groups here at this event. Immediately following my presentation we will have Bill Anderson and Brad Anderson — they’re not brothers — Bill Anderson and Brad Anderson come on stage and talk to you in more detail about SMS version 4 and very excited about your feedback after that presentation.
The Roadmap in Perspective
So putting this roadmap together, putting it in perspective, essentially we’re seeing waves of delivery of the System Center family of products. There are systems management products that we talked about, SMS, MOM, Data Protection Manager; there are some core technologies that we deliver in wave one of System Center that I didn’t specifically address. Visual Studio 2005 that will be available in the latter half of this year will deliver SDM to market. This is a very pinnacle release of Visual Studio that’s actually going to get the first version of SDM to the marketplace.
And Windows Server 2003 R2, which will also be available in the second half of this calendar year, will essentially bring WS-Management to market. This will be the first release of Windows from Microsoft getting WS-Management to the marketplace and essentially paving the way to future waves of System Center, taking advantage of both SDM and WS-Management.
In wave two you will see all the innovations that we talked about in MOM, SMS, new version of Reporting Manager, new version of System Center Capacity Manager, essentially extending the set of scenarios of that product from just Exchange and MOM to a complete range of technologies from Microsoft and Microsoft partners.
We’re very excited about it, very excited about this roadmap and really looking forward to speaking with you in the next 12 to 24 months, reporting on our progress along the way.
And System Center wave 3 is something that we’re targeting towards 2008, 2009, something that we’re not talking about at this conference, but looking forward to talking to you in more detail in the not-so-distant future.
Now, I mentioned that we’re really thinking about this roadmap and we’re really thinking about delivery on this roadmap from the methodology from the Microsoft Operations Framework standpoint, and this is how this roadmap maps into MOF quadrants.
Naturally Systems Management Server is prime technology for changing configuration manager, Microsoft Operations Manager and data protection manager fits in the operational quadrant where these products are used to run their environment day-to-day.
In the area of supporting and servicing there are some very interesting things that are happening. First of all, we’re working very closely with partners, with partners who deliver IT service desk technology to the market. One partner I would highlight is a relatively small French company called PS’Soft that in our view has very innovative technology in both IT assets manager and integrated service desk and they partner with us, they partner with SMS to deliver this product to the market.
There are also a couple of technologies from Microsoft that are not widely known in the industry and I thought it important to highlight them. First of all, it’s Microsoft CRM solution that comes from our Microsoft Business Solutions product family and that technology primarily addresses the scenario of the outfacing help desk, basically customer relationship management solutions. However, in the last six months we saw some of our customers take advantage of MS CRM, integrate it with Microsoft Operations Manager and start using MS CRM as IT service desk for their IT organizations. It’s a very interesting step, very interesting development and we’re obviously going to follow that.
And finally we talked about Visual Studio 2005 to be available the latter part of this year. Within the Visual Studio 2005 family we’re delivering Visual Studio Team System, which is a much broader view on individual personas and individual connections and individual servicing needs and support needs within the development domain, where it’s not just developers who we target anymore but it’s also test, architect and support person that now can cooperate through this Visual Studio Team System. So yet again this technology is available through the Visual Studio beta 2 program and I would certainly encourage you to take a look at it and see if it can fit your needs.
And finally in the area of optimization and planning, which to me is the most exciting, the most proactive part of the Microsoft Operations Framework, System Center Reporting Manager with all the wealth of business analytic tools coming with that product and System Center Capacity Manager will greatly help you implement those technologies and deliver on the methodology of the Microsoft Operations Framework.
So technology really feels the quadrant of Microsoft Operations Framework, this technology all comes with well defined methodology, with all defined solution accelerators, with best practices and prescriptive guides, giving you advantage to implement it in your environment and start taking advantage of the technologies much quicker.
Why Microsoft for Enterprise Systems Management
Let’s move to a Microsoft discussion and let’s talk about why we think Microsoft is uniquely qualified to be your partner in the enterprise systems management.
The first very important area I want to address is integrated innovation. In my three years working with Microsoft there has been a dramatic shift towards the group working together, towards the group integrating their product plans, integrating their test plans with the goal that our customers need tightly integrated solutions, our customers need tightly integrated environments and that’s exactly what we’re set to deliver from Microsoft. Our internal policies, our internal ways of doing things have changed; we’re much more of an integrated company delivering integrated solutions, delivering Windows Server System tightly linked with the rest of what we deliver to you.
There are some concrete steps that we took in that area in the last year. We launched Windows Server System common engineering criteria, which is basically a set of guidelines that we deploy internally on how we build products and what kind of rules and guidelines this product ought to follow. As you can see, we demand that every server-based technology that we ship comes with Microsoft management packs, we demand that this technology can be serviced through Windows Update, we demand that this technology comes with well-defined architectural guides and manageability guides. We follow error reporting, we ensure high availability, we ensure support of virtual server, security, 64-bit, a variety of guidelines to come to developers inside Microsoft, essentially telling them what needs to be done before the product comes to market.
We’re also being very public with what we do here. Common engineering criteria is published on the Microsoft Web site and you can take a look at how our products today comply with these engineering criteria.
Now, this policy and this practice that we employ internally in Microsoft in putting manageability and best practices into everything as we do is another very important aspect of Dynamic Systems Initiative. Technology is just one aspect of DSI; methodology, guidelines and change of the way people traditionally do things is what needs to happen. Developers need to be thinking about operations, developers need to be thinking how the technology would actually run in the end user’s environment. And this is the goal of common engineering criteria, to educate developers inside Microsoft about this need, which would enable us to get DSI embedded into all of Microsoft’s products, but it’s also about us being very public about it and it’s also about us opening common engineering criteria to the rest of the industry and essentially bringing the entire ecosystem of partners with that to make sure that our hardware partners, our solution partners and our services partners follow these common engineering criteria and deliver Dynamic Systems Initiative jointly with us to you, to the industry, essentially positioning Microsoft as a unique vendor to solve your enterprise management needs.
So we talked about progress, we talked about progress of what we delivered in the last 12 months. We talked about Dynamic Systems Initiative, we talked about the continuous investment that we make in defining strategy and concrete steps that we’re making in delivering this technology to you through our roadmap. And we talked about the 24 month roadmap and how we’re trying to follow your feedback and be much more predictable on what we deliver and when and this exciting roadmap I believe will deliver to your needs in a very predictable fashion. And finally we talked about Microsoft and the unique qualifications of Microsoft in becoming your partner in enterprise systems management.
This is a very exciting event, this is a very exciting event for you to explore technologies that we deliver to you, this is a very exciting event for you to exchange ideas, this is a very exciting event for you to experience technologies that we talk about in the hands-on labs. And those labs are open, the labs are fully staffed and I’d fully encourage you to go and play with these products, see what you think about it, give us your feedback, help us deliver better solutions to you.
MMS 2005 is just one step in this journey of delivering Dynamic Systems Initiative and delivering our technologies to you. I’m really looking forward to continuing this journey and I’m actually looking forward to the Microsoft Management Summit 2006 and I want to unveil Microsoft Management Summit 2006 to you. We’re changing the venue, we’re moving to San Diego. It’s not too early to put it — (applause.) Thank you. It’s not too early to put it on your agenda. I’m really looking forward to it, I’m really looking forward to delivering to your needs in systems management and DSI and I really thank you for joining us for this exciting event.
Enjoy the conference. Thank you very much. (Applause.)