1 David Chappell March 2012 WINDOWS AZURE AND WINDOWS HPC SERVER HIGH-PERFORMANCE COMPUTING IN THE CLOUD Sponsored by Microsoft Corporation Copyright 2012 Chappell & Associates
2 Contents High-Performance Computing and the Cloud... 3 Technology Basics... 4 Windows Azure...4 Windows HPC Server...6 Using Windows HPC Server with Windows Azure... 9 Using On-Premises Computing and Cloud Computing Together...9 Using Cloud Computing Exclusively...11 Scenarios Embarrassingly Parallel Applications...13 Excel Applications...14 MPI Applications...16 An ISV Scenario: Using the Windows Azure HPC Scheduler...17 Conclusion About the Author
3 High-Performance Computing and the Cloud The essence of high-performance computing (HPC) is processing power. Whether implemented using traditional supercomputers or more modern compute clusters, HPC requires having lots of available computing resources. Windows HPC Server 2008 R2 lets organizations provide these resources. By letting machines be managed as a cluster, then providing tools for deploying and running HPC applications on that cluster, the product makes Windows a foundation for high-performance computing. A variety of organizations today use Windows HPC Server clusters running in their own data centers to solve a range of hard problems. But with the rise of cloud computing, the world of HPC is changing. Why not take advantage of the massive data centers now available in the cloud? For example, Microsoft s Windows Azure provides on-demand access to an abundance of virtual machines (VMs) and acres of cheap storage, letting you pay only for the resources you use. This is appealing for applications that need lots of processing power. The potential benefits for HPC are obvious. Rather than relying solely on your own on-premises computing resources, using the cloud gives you access to more compute power when you need it. For example, suppose your on-premises systems are sufficient for most but not all of your organization s workloads. Or suppose one of your HPC applications sometimes needs more processing power. Rather than buy more machines, you can instead rely on a cloud data center to provide extra computing resources on demand. Depending on the kinds of applications your organization runs, it might even be feasible to eventually move more and more of your HPC work into the cloud. This allows tilting HPC costs away from capital expense and toward operating expense, something that s attractive to many organizations. Yet relying solely on the cloud for HPC might not work for many organizations. The challenges include the following: Like other applications, HPC applications (often called jobs) sometimes work on sensitive data. Legal, regulatory, or other limitations might make it impossible to store or process that data in the cloud. A significant number of HPC jobs rely on applications provided by independent software vendors (ISVs). Because many of these ISVs have yet to make their applications available in the cloud, a significant number of HPC jobs can t use cloud platforms today. Jobs that need lots of computing power often rely on large amounts of data. Moving all of that data to a cloud data center can be problematic. While the rise of cloud computing will surely have a big impact on the HPC world, the reality is that on-premises HPC computing isn t going away. Instead, providing a way to combine the two approaches on-premises and cloud makes sense. This is exactly what Microsoft has done with Windows HPC Server and Windows Azure, supporting all three possible combinations: applications that run entirely on-premises, applications that run entirely in the cloud, and applications that are spread across both. This paper describes how Windows HPC Server R2 Service Pack (SP) 2 and Windows Azure can work together. After a brief tutorial on each one, it walks through the options for combining them. The goal is to provide an architectural overview of what this technology is and why users of HPC applications should care. 3
4 Technology Basics Understanding how Windows HPC Server and Windows Azure can be combined for high-performance computing requires understanding the basics of both technologies. This section walks through the fundamentals of each one, starting with Windows Azure. Windows Azure Windows Azure is a public cloud platform providing Internet-accessible computers running in Microsoft data centers. Its customers run applications and store data on these machines, paying Microsoft for the computing and storage resources they use. Figure 1 shows the core Windows Azure components. Figure 1: A Windows Azure data center runs applications and stores data. As the figure shows, every Windows Azure application runs on computers in a Windows Azure data center. Each of these applications runs in one or more instances, which are really VMs. For storage, Windows Azure provides three main options: SQL Azure, a managed service for relational data based on SQL Server. Tables, providing key-based scalable storage for simple typed data. (Don t be confused despite its name, this option doesn t provide relational storage.) Blobs, offering unstructured storage for binary data. Windows Azure also provides built-in management services, including things such as automatically deploying patches and other updates to the operating systems running in virtual and physical machines. 4
5 Whatever its purpose, a Windows Azure application is always implemented as one or more roles. Three role types are supported today: Web roles, which are typically used for code that accepts HTTP requests via Internet Information Services (IIS). Worker roles, providing a more general option that can be used for various kinds of processing. VM roles, which allow a Windows Azure user to upload a VM image as a Windows Server 2008 R2 VHD, then have Windows Azure execute that image. Whatever roles an application uses, Windows Azure generally requires it to run at least two instances of each one. Depending on the work it needs to do, an application might run two, ten, or fifty instances of a role. Figure 2 illustrates this idea. Figure 2: A Windows Azure application consists of some combination of role instances, each running as a separate virtual machine. In this example, the Windows Azure application is implemented using one Web role, one Worker role, and one VM role. It s currently running two instances of each role, but changing this is straightforward. A developer or administrator can tell Windows Azure to increase or decrease the number of instances, or the application can do this itself. As the figure shows, Windows Azure typically assigns each instance in the application to a different computer. This helps make applications more fault-tolerant, since a single hardware failure won t take down the entire application. The platform also monitors each running application. If an instance fails, Windows Azure starts a new instance of the same role within a specified time defined by the platform s service level agreement (SLA). And 5
6 because each instance has one or more dedicated processer cores, an application s performance is predictable it doesn t share those cores with other applications. Each instance runs a version of Windows Server, and so the software it runs can be created using the.net Framework and a language such as C#. Those applications can also be built using raw C++, however, or other technologies such as PHP and Java. And while Visual Studio is today s most commonly used tool for creating Windows Azure applications, it s not the only choice developers are free to use whatever tools they like. However it s built, a Windows Azure application can access data stored in any of the three services described earlier: SQL Azure, tables, and blobs. An application is also free to access on-premises data within an organization there s no requirement that a cloud application use only data stored in the cloud. Making good decisions about using Windows Azure requires a basic understanding how it s priced. (For current prices and more detail, see here.) Here s a summary of the main things to know: For compute, customers are charged per instance (i.e., per VM) per hour. The price varies based on the instance s size, with options ranging from one to eight cores. Note that this charge isn t usage-based; an instance that sits unused for an hour will incur the same charge as an instance that spends an hour doing intensive calculations. For storage, data kept in Windows Azure blobs and tables is charged per gigabyte per month. Customers also pay a small amount for operations on this data, such as reads and writes. For bandwidth, customers pay for each gigabyte of data moved out of a Windows Azure datacenter. (Moving data into a Windows Azure datacenter is free.) There s no bandwidth charge for Windows Azure applications that access blobs, tables, or SQL Azure data stored in the same datacenter, however. Cloud platforms offer a new approach to deploying, using, and paying for computing resources. With Windows Azure, Microsoft provides a way for Windows-oriented developers to embrace this new style of computing. Windows HPC Server The essence of high-performance computing is dividing an application into separate components, then running those components simultaneously on multiple machines. With Windows HPC Server, this means spreading the application s logic across multiple compute nodes, each a physical computer or VM running Windows. The application might run the same logic on every compute node, with each instance processing different data, or it might run different logic on different nodes. To assign the logic of a Windows HPC application to compute nodes, Windows HPC Server provides an HPC job scheduler. This scheduler runs on its own dedicated machine called a head node. Figure 3 shows how this looks. 6
7 Figure 3: An HPC job scheduler running on a head node distributes the logic of a Windows HPC application across a group of compute nodes. As the figure shows, a user typically submits a Windows HPC job from a workstation to the head node. The HPC job scheduler parcels the job s logic across the compute nodes that this application has access to. Traditionally, a computing cluster contained only on-premises servers. With Windows HPC Server today, however, a cluster can include any combination of three options: Desktop workstations running Windows 7. On-premises servers running either Windows HPC Server 2008 R2 or its predecessor, Windows HPC Server These can be physical or virtual servers. Windows Azure instances, including both Worker roles and VM roles. (Web roles aren t typically used with Windows HPC Server.) Assigning work to compute nodes isn t a simple task, and so the HPC job scheduler is a sophisticated piece of software. Among other things, it runs applications based on their priority, allowing new high-priority jobs to jump to the head of the line. It also lets the user submitting a job specify what kind of resources the job needs, then places the job appropriately. Using this infrastructure, Windows HPC Server can run several kinds of applications. The options today include the following: Applications where components running on different computers in the cluster interact with one another while the application is running. Because this interaction typically happens via a technology called the Message 7
8 Passing Interface (MPI), these are commonly known as MPI applications. Many scientific and technical problems can be addressed with MPI applications, including those that require simulating physical processes such as car crashes and nuclear reactions. Applications where components running on different computers in the cluster do not interact with one another while the application is running. Because applications like these are so easy to run in parallel, they re referred to as embarrassingly parallel applications. This kind of application is especially common in financial services, addressing problems such as evaluating the risk of an investment portfolio. Excel applications that offload a workbook s calculations to a cluster, speeding up work that would take much longer to complete on a single machine. Excel workbooks are used for a wide range of problems, many of which perform time-consuming calculations that can benefit from running on a cluster. Much of what Windows HPC Server offers is focused on creating and running applications. Yet managing a cluster and the applications that run on it is also a non-trivial task. For example, an administrator must create the cluster, then determine things such as which compute nodes are available for each job. An application might have a specific set assigned to it, or it might be able to draw compute nodes from a generally available pool. To make life easier for cluster administrators, Windows HPC Server includes HPC Cluster Manager. Figure 4 shows an example of how this tool looks. Figure 4: HPC Cluster Manager lets a cluster administrator add nodes to a cluster, monitor and manage jobs running on the cluster, and much more. 8
9 As the figure suggests, an administrator uses this tool to accomplish a variety of tasks: managing a cluster s nodes, managing jobs running on that cluster, running diagnostic tests on the cluster, creating charts and reports about the cluster, and more. HPC Cluster Manager also provides graphical views, including a heat map showing the utilization of each node in a cluster. And even though a cluster can contain any combination of desktop workstations, on-premises servers, and Windows Azure instances, HPC Cluster Manager lets a cluster administrator work with all three in a consistent way. High-performance computing isn t an especially simple area. Building and running HPC applications requires specialized software, as does managing the clusters on which they run. Windows HPC Server aims at providing a cohesive solution for all of these problems. Using Windows HPC Server with Windows Azure HPC depends on processing power. Windows Azure provides a vast amount of processing power on demand. The two are clearly a natural combination. But for all of the reasons described earlier, the rise of the cloud doesn t signal the end of on-premises HPC computing. Instead, the on-premises and cloud worlds can work together, as the marriage of Windows HPC Server and Windows Azure demonstrates. This combination supports two approaches today: Keeping the head node and some compute nodes on premises, with extra compute nodes in the cloud. Keeping only the head node on premises, with all compute nodes in the cloud. This section looks at both options. Using On-Premises Computing and Cloud Computing Together From an architectural perspective, running a Windows HPC application that uses both on-premises compute nodes and Windows Azure instances is straightforward. Figure 5 shows how this looks. 9
10 Figure 5: A Windows HPC application can use both on-premises compute nodes and Azure nodes. The head node remains on premises, and a user submits the HPC job in the usual way. The HPC job scheduler on the head node then parcels out the application s logic across the available compute resources. In this case, those resources include both on-premises compute nodes and Azure nodes, which are Worker role and/or VM role instances running in a Windows Azure data center. To a user, submitting a Windows HPC job that runs partly on-premises and partly in the cloud looks just as it always does there s no difference from submitting one that runs entirely on-premises. For the Windows HPC administrator, however, some extra work is required to make this possible. While the admin still uses HPC Cluster Manager and other familiar tools, she must set up a Windows Azure account and configure the Azure nodes. The administrator also specifies which jobs are allowed to use which compute resources, whether those resources are compute nodes in the on-premises cluster or Azure nodes in the cloud. Azure nodes aren t created on demand when a job is submitted. Instead, the administrator explicitly allocates these VMs, then shuts them down when they re no longer needed. For example, an administrator might create 25 Azure nodes at 9 am each day, then take them down at 6 pm. This can be done manually through HPC Cluster Manager or programmatically via a script. Using Azure nodes costs money each one is a Windows Azure instance and so Windows HPC Server gives an administrator control over when those instances are created and how long they run. Although it s not shown in the diagram, this application almost certainly reads data in and writes data out. That data might be stored on premises, with the local compute nodes having direct access while the Azure nodes access it across the Internet. The application s data might also be stored in the cloud using, say, blobs in Windows Azure storage, or it might be divided between on-premises storage and cloud storage. Windows HPC Server doesn t mandate where data is stored or how it s accessed it s up to the creator of each application to make this decision. 10
11 What kind of applications would benefit from combining on-premises compute nodes and cloud-based Azure nodes? When would this approach be useful? Here are some examples: An application that sometimes needs extra compute power could use Windows Azure to provide this on demand. Rather than investing in more machines on-premises, an organization can use the cloud to get the resources it needs at a lower cost. An organization that periodically needs extra HPC compute nodes might use Windows Azure to provide them. Think of a financial services firm, for example, that runs a group of CPU-intensive jobs once a week. Rather than expand its existing on-premises cluster, then watch those new machines sit idle most of the time, this firm could instead allocate Azure nodes only when needed. An organization might be running out of room for its on-premises cluster. Rather than expand by adding more physical machines to a crowded data center, they could instead choose to expand into the cloud by using Azure nodes. As mentioned earlier, this also avoids the capital expense of buying new hardware, replacing it with the operating expense of using resources in the cloud. And because Windows Azure automates routine tasks like patching Windows, the administrative costs of an HPC environment can also be reduced. Whatever the reason, developers creating HPC applications that run across Windows HPC Server and Windows Azure see a familiar world. They can still use Visual Studio (or other Windows development tools) and rely on the product s standard profiling and debugging support. The big change is that their applications can now have access to significantly more computing power. Using Cloud Computing Exclusively If putting some compute nodes in Windows Azure makes sense and it often does why not put all of them in the cloud? Windows HPC Server also supports this option, as Figure 6 shows. 11
12 Figure 6: A Windows HPC application can rely entirely on Azure nodes. In this case, only the head node remains on premises. All of the processing is done in Azure nodes running in a Windows Azure datacenter. Once again, the user submitting a Windows HPC job sees no difference the scheduler transparently spreads the application s logic across the Azure nodes. Putting all of an application s computing in the cloud can have some clear benefits. They include these: An organization can run HPC jobs without buying and managing its own cluster. Maybe the utilization of an onpremises cluster wouldn t be high enough to justify the expense, or perhaps a firm needs HPC capabilities only once a month. Whatever the situation, having just the head node on premises with all computing done on Azure nodes can make economic sense. The cost of entry to HPC can be significantly lower than with an on-premises cluster. Rather than spending a chunk of money up front on hardware, an organization can instead pay only for the HPC resources it needs when it uses them. Compared to dividing an HPC application between on-premises resources and Windows Azure, running the application entirely in the cloud can make data access easier. Rather than splitting the data between the two worlds or making one part of the application access data remotely, all of the application s data can be stored in the cloud. For some organizations, the SLA provided by Windows Azure might provide more reliability than their own data center. And since Windows Azure automatically handles updates to system software, the instability these can cause is minimized. Whether you choose to split an HPC application between on-premises resources and the cloud or to run that application entirely in the cloud, the potential benefits of combining HPC and cloud computing are evident. While 12
13 it s not right for every situation, expect to see an increasing amount of HPC workloads taking advantage of what the cloud has to offer. Scenarios To understand how Windows HPC Server and Windows Azure can work together, it s useful to look at each of the three kinds of applications this combination supports today: embarrassingly parallel jobs, Excel jobs, and MPI jobs. What follows takes a brief look at each one. Embarrassingly Parallel Applications In an embarrassingly parallel job, the logic running on each compute node can work in isolation there s no need for the components to interact while the job runs. Windows HPC Server supports two approaches to creating this kind of application: service-oriented architecture (SOA) applications and parametric sweep applications. As the name suggests, the compute logic in a SOA application is implemented as a service. With Windows HPC Server, each service is implemented using Windows Communication Foundation (WCF), allowing the service to expose its operations via SOAP or REST or something else. This approach lets users interact with the job while it s running. In a parametric sweep application, the compute logic is implemented as an ordinary executable. Rather than exposing services that let a user interact with the job while it s running, the submitter instead can pass different initial parameters to the executable on each compute node when the job starts running. This lets each executable apply the same logic to different data. And while parametric sweep applications can t interact with users, they also don t require using WCF. This can make them easier for scientists and other non-professional developers to create. To see how this kind of solution works with Windows HPC Server and Windows Azure, suppose you need to write a risk analysis application using a Monte Carlo simulation. This application is likely to be embarrassingly parallel it can easily be broken up into chunks, each of which works independently. Each chunk runs on its own compute node, and each one can read and write data from wherever the developer chooses, such as a shared file system. With this kind of simulation, running on more nodes typically leads to better results. Suppose that running the application using only on-premises compute nodes is usually sufficient. In some cases, however, greater accuracy might be required, something that using more compute nodes can provide. Rather than investing in new computers that won t be fully used, the organization running this application might instead choose to rely on Windows Azure to do the extra work. Figure 7 shows how this looks. 13
14 Figure 7: An embarrassingly parallel application might have its work divided between an on-premises data center and Windows Azure. As the figure shows, the user submits the job as always (step 1). (Although it s not shown here, this relies on the head node as described earlier.) Once the job is executing, the user can interact with it if it s a SOA application (step 2). In any case, the job will eventually return results to the user (step 3). The Azure nodes used here might be either Worker role instances or VM role instances. If it s a SOA job, its creator probably chose Worker roles, since they re simpler to use and they support WCF. Some existing applications can be challenging to run in Worker roles, however. If this is a parametric sweep job, for instance, it might rely on existing Windows software that can t easily be installed in a Worker role. In this case, the application s creator can instead generate a Windows Server 2008 R2 VHD that contains the application s software, then upload this VHD into Worker roles. Windows HPC Server makes things simpler by providing a graphical tool to do this. Figure 7 shows an embarrassingly parallel application spread across both on-premises compute nodes and Azure nodes. It s also possible, of course, to run the application entirely in the cloud or entirely on-premises. It s up to the user to decide which option makes the most sense. Excel Applications Excel can use a Windows HPC cluster in several different ways. The options are: Using Excel as the client for a SOA job. With Visual Studio Tools for Office, a developer can write a custom SOA application that gets executed from an Excel workbook. 14
15 Running Excel user-defined functions (UDFs) on a cluster. UDFs allow creating custom code that can be called from spreadsheet cells just as if they were built-in Excel functions. The code that implements a UDF can run on a Windows HPC cluster. Running Excel workbooks on a cluster. An organization can install Excel on multiple compute nodes in a cluster, then have each copy run the same workbook simultaneously with different data. With any of these options, the cluster can be entirely on-premises, entirely in the cloud, or divided between the two. For example, suppose a financial analyst creates an Excel workbook that models the behavior of a group of transactions. If the number of transactions is usually small, the analyst might normally run the workbook on his own desktop workstation. In some cases, however, he might need to work with a much larger number of transactions. To do this effectively, he can execute the workbook on Windows Azure instead. Figure 8 shows how this looks. Figure 8: An Excel application can perform workbook calculations on Windows Azure. In this scenario, the user sends the workbook to Windows Azure as usual (step 1). Because Excel is installed on multiple Azure nodes, the workbook s calculations can run simultaneously on each one, with every copy processing different data. When the job completes, the result is returned to the user (step 2). Running Excel on Windows Azure in this way requires using VM roles Excel can t be installed in a Worker role. Before this job is run, someone would need to create a Windows Server 2008 R2 VHD that included Excel, then upload this image to Windows Azure. Most likely, an administrator would handle this, letting the financial analyst remain unaware of these technical details. 15
16 MPI Applications With both embarrassingly parallel applications and Excel applications, each chunk of logic works independently the chunks don t need to interact while the job runs. With MPI applications, however, the chunks communicate to exchange intermediate results or for other reasons. Yet these jobs can still run on Windows Azure, as Figure 9 shows. Figure 9: An MPI application can run entirely on Windows Azure. MPI applications are typically batch jobs. As shown here, the user submits the job (step 1), then gets the job s result returned to her (step 2). As in the earlier cases, an MPI application can run entirely on Windows Azure, as in this example, entirely on-premises, or be divided between the two. An MPI application can use either Worker roles or VM roles. VM roles require slightly more administration, but they can also make it easier to run some existing software, as described earlier. If an organization wishes to run a purchased MPI application on Windows Azure, for example, using VM roles might make this simpler, minimizing the need for Windows Azure-specific modifications by the application s creator. It s worth pointing out that because the components in an MPI job must interact while the job runs, on-premises clusters that support this kind of application commonly connect their servers with high-bandwidth networks such as InfiniBand. Windows Azure datacenters don t provide these high-speed connections between machines, and so MPI jobs that depend on very fast communication might not perform well in the cloud. Similarly, dividing an MPI job s processing between on-premises systems and Windows Azure can be problematic for the same reason: The communication between these two worlds might not be fast enough. Still, a significant fraction of MPI jobs can run quite well in the cloud, and it s a good option in some situations. 16
17 An ISV Scenario: Using the Windows Azure HPC Scheduler One more option for using HPC with Windows Azure relies on a technology that hasn t yet been described: the Windows Azure HPC Scheduler. This component is most useful to independent software vendors (ISVs), and so here s a typical situation in which it might be used. Suppose an ISV wishes to create a multi-tenant HPC application running on Windows Azure, i.e., one that can be used by many different customers. This application might do risk calculations or portfolio modeling or something else. Whatever it does, it breaks work down into a number of parts, then runs those parts in parallel using many Windows Azure instances. As with any HPC application, some kind of scheduler is required to decide how to spread work across these instances. The ISV is free to create this scheduler itself, of course, but to make life easier, Microsoft provides the Windows Azure HPC Scheduler. Figure 10 shows how this looks. Figure 10: An ISV can incorporate the Windows Azure HPC Scheduler into an HPC application that runs on Windows Azure. The Windows Azure HPC Scheduler implements only the scheduling aspects of the HPC job scheduler it doesn t provide deployment and so it s not a standalone head node in the cloud. Some of the management tools provided with the Windows HPC Server head node can work with the Windows Azure HPC Scheduler, however, such as examining job queues. Still, there s no requirement that a Windows Azure application using this cloud scheduler also use Windows HPC Server; the two technologies aren t dependent on each other. The Windows Azure HPC Scheduler supports all three kinds of HPC applications: embarrassingly parallel, Excel, and MPI. It uses SQL Azure for its own storage, but applications built with it can use any storage mechanism they 17
18 choose. And there s no extra charge for including it in custom software. A running application built with the Windows Azure HPC Scheduler just pays the normal Windows Azure usage fees as always. Conclusion High-performance computing is entering a new era. The enormous scale and low cost of cloud computing resources is sure to change how and where HPC applications run. Ignoring this change isn t an option. Yet the rise of the cloud doesn t imply the death of today s clusters. Data issues, networking limitations, and other concerns mean that some HPC jobs remain best suited for on-premises computing. This reality is why Windows HPC Server supports all three possibilities: Running applications entirely in on-premises systems. Running applications entirely in the cloud. Running applications across both on-premises systems and the cloud. HPC applications can help solve some of the world s most important problems. Cloud computing is among the most important technology trends today. It shouldn t be surprising that the two are a natural combination. About the Author David Chappell is Principal of Chappell & Associates (www.davidchappell.com) in San Francisco, California. Through his speaking, writing, and consulting, he helps people around the world understand, use, and make better decisions about new technologies. 18
David Chappell June 2011 CLUSTER COMPUTING TODAY WHAT S CHANGED AND WHY IT MATTERS Sponsored by Microsoft Corporation Copyright 2011 Chappell & Associates One way to make an application run faster is to
THE WINDOWS AZURE PROGRAMMING MODEL DAVID CHAPPELL OCTOBER 2010 SPONSORED BY MICROSOFT CORPORATION CONTENTS Why Create a New Programming Model?... 3 The Three Rules of the Windows Azure Programming Model...
WINDOWS AZURE EXECUTION MODELS Windows Azure provides three different execution models for running applications: Virtual Machines, Web Sites, and Cloud Services. Each one provides a different set of services,
David Chappell October 2012 WINDOWS AZURE DATA MANAGEMENT CHOOSING THE RIGHT TECHNOLOGY Sponsored by Microsoft Corporation Copyright 2012 Chappell & Associates Contents Windows Azure Data Management: A
David Chappell December 2011 WHAT IS AN APPLICATION PLATFORM? Sponsored by Microsoft Corporation Copyright 2011 Chappell & Associates Just about every application today relies on other software: operating
WINDOWS AZURE AND ISVS A GUIDE FOR DECISION MAKERS DAVID CHAPPELL JULY 2009 SPONSORED BY MICROSOFT CORPORATION CONTENTS ISVs and Cloud Computing... 2 A Brief Overview of Windows Azure... 3 Technology...
WINDOWS AZURE DATA MANAGEMENT AND BUSINESS ANALYTICS Managing and analyzing data in the cloud is just as important as it is anywhere else. To let you do this, Windows Azure provides a range of technologies
WINDOWS AZURE NETWORKING The easiest way to connect to Windows Azure applications and data is through an ordinary Internet connection. But this simple solution isn t always the best approach. Windows Azure
SELLING SHAREPOINT ENGAGEMENTS IN THE CLOUD ERA A GUIDE FOR MICROSOFT SI PARTNERS Sponsored by Microsoft Corporation 1/ Selling SharePoint Online 2/ Selling SharePoint Farms on Windows Azure 3/ Selling
Windows HPC Server 2008 R2 Service Pack 3 (V3 SP3) Greg Burgess, Principal Development Manager Windows Azure High Performance Computing Microsoft Corporation HPC Server Components Job Scheduler Distributed
INTRODUCING WINDOWS AZURE Windows Azure is Microsoft s application platform for the public cloud. You can use this platform in many different ways. For instance, you can use Windows Azure to build a web
David Chappell ADOPTING MICROSOFT AZURE A GUIDE FOR IT LEADERS Sponsored by Microsoft Corporation Copyright 2014 Chappell & Associates Contents Public Cloud Platforms: The Future of Enterprise Computing...
GIS IN THE CLOUD THE ESRI EXAMPLE DAVID CHAPPELL SEPTEMBER 2010 SPONSORED BY ESRI CONTENTS Contents... 2 Cloud Computing Basics... 3 Cloud Applications and Cloud Platforms... 3 An Example Cloud Platform:
INTRODUCING WINDOWS AZURE DAVID CHAPPELL OCTOBER 2010 SPONSORED BY MICROSOFT CORPORATION CONTENTS An Overview of Windows Azure... 2 Compute... 4 Storage... 5 Fabric Controller... 7 Content Delivery Network...
THE BENEFITS AND RISKS OF CLOUD PLATFORMS A GUIDE FOR BUSINESS LEADERS DAVID CHAPPELL JANUARY 2011 SPONSORED BY MICROSOFT CORPORATION Cloud platforms are a fundamental part of the move to cloud computing.
David Chappell SELLING PROJECTS ON THE MICROSOFT BUSINESS ANALYTICS PLATFORM A PERSPECTIVE FOR SYSTEMS INTEGRATORS Sponsored by Microsoft Corporation Copyright 2014 Chappell & Associates Contents Business
David Chappell INTRODUCING AZURE SEARCH Sponsored by Microsoft Corporation Copyright 2015 Chappell & Associates Contents Understanding Azure Search... 3 What Azure Search Provides...3 What s Required to
David Chappell OPERATIONAL SCENARIOS USING THE MICROSOFT DATA PLATFORM A GUIDE FOR IT LEADERS Sponsored by Microsoft Corporation Copyright 2016 Chappell & Associates Contents Microsoft s Data Platform:
IT as a Service Transforming IT with the Windows Azure Platform November 2010 Version 1.0 11/9/2010 Contents Understanding IT as a Service... 1 Realizing IT as a Service: The Importance of PaaS... 4 What
A SHORT INTRODUCTION TO CLOUD PLATFORMS AN ENTERPRISE-ORIENTED VIEW DAVID CHAPPELL AUGUST 2008 SPONSORED BY MICROSOFT CORPORATION COPYRIGHT 2008 CHAPPELL & ASSOCIATES CONTENTS Defining Terms: What is a
INTEGRATION SOFTWARE: BUILD OR BUY? DAVID CHAPPELL MARCH 2010 SPONSORED BY MICROSOFT CORPORATION No modern organization uses only one application. A single software solution just isn t enough to address
PROVIDING SINGLE SIGN-ON TO AMAZON EC2 APPLICATIONS FROM AN ON-PREMISES WINDOWS DOMAIN CONNECTING TO THE CLOUD DAVID CHAPPELL DECEMBER 2009 SPONSORED BY AMAZON AND MICROSOFT CORPORATION CONTENTS The Challenge:
INTRODUCING WINDOWS AZURE DAVID CHAPPELL MARCH 2009 SPONSORED BY MICROSOFT CORPORATION CONTENTS An Overview of Windows Azure... 2 The Compute Service... 3 The Storage Service... 5 The Fabric... 7 Using
What is Application Lifecycle Management? David Chappell Sponsored by Microsoft Corporation Copyright 2014 Chappell & Associates Defining application lifecycle management (ALM) isn t easy. Different people
David Chappell Introducing DocumentDB A NoSQL Database for Microsoft Azure Sponsored by Microsoft Corporation Copyright 2014 Chappell & Associates Contents Why DocumentDB?... 3 The DocumentDB Data Model...
INTRODUCING WINDOWS AZURE DAVID CHAPPELL DECEMBER 2009 SPONSORED BY MICROSOFT CORPORATION CONTENTS An Overview of Windows Azure... 2 The Compute Service... 3 The Storage Service... 5 The Fabric... 7 Using
INTRODUCING THE WINDOWS AZURE PLATFORM AN EARLY LOOK AT WINDOWS AZURE, SQL AZURE, AND.NET SERVICES DAVID CHAPPELL AUGUST 2009 SPONSORED BY MICROSOFT CORPORATION CONTENTS An Overview of the Windows Azure
David Chappell Microsoft Azure Data Technologies: An Overview Sponsored by Microsoft Corporation Copyright 2014 Chappell & Associates Contents Blobs... 3 Running a DBMS in a Virtual Machine... 4 SQL Database...
David Chappell Understanding NoSQL on Microsoft Azure Sponsored by Microsoft Corporation Copyright 2014 Chappell & Associates Contents Data on Azure: The Big Picture... 3 Relational Technology: A Quick
INTRODUCING BLUE PRISM AUTOMATING BUSINESS PROCESSES WITH PRESENTATION INTEGRATION DAVID CHAPPELL MARCH 2010 SPONSORED BY BLUE PRISM CONTENTS Business Process Automation with Blue Prism... 3 The Role of
INTRODUCING THE WINDOWS AZURE PLATFORM DAVID CHAPPELL OCTOBER 2010 SPONSORED BY MICROSOFT CORPORATION CONTENTS An Overview of the Windows Azure Platform... 3 Windows Azure... 4 SQL Azure... 6 Windows Azure
Selling Windows Azure Projects CUSTOM APPLICATIONS A GUIDE FOR MICROSOFT SI PARTNERS Sponsored by Microsoft Corporation 1/ Why Should You Sell Custom Application Projects on Windows Azure? 2/ How Can You
Microsoft HPC V 1.0 José M. Cámara (firstname.lastname@example.org) Introduction Microsoft High Performance Computing Package addresses computing power from a rather different approach. It is mainly focused on commodity
Cloud Platforms in the Enterprise A Guide for IT Leaders @DChappellAssoc Copyright 2014 Chappell & Associates The Three Most Important IT Events In the last decade Initial pubic offering of Salesforce.com,
MICROSOFT DYNAMICS CRM Comparing the xrm Application Framework and Force.com: A Guide for Technical Decision Makers January 2011 CONTENTS Foundations for Business Applications: xrm and Force.com... 3 Why
Windows Azure platform What is in it for you? Dominick Baier (email@example.com) Christian Weyer (firstname.lastname@example.org Objectives Motivation Status quo Cloud Computing Windows Azure platform Windows Azure
WHAT IS APPLICATION LIFECYCLE MANAGEMENT? DAVID CHAPPELL DECEMBER 2008 SPONSORED BY MICROSOFT CORPORATION COPYRIGHT 2008 CHAPPELL & ASSOCIATES Defining application lifecycle management (ALM) isn t easy.
Presented at the COMSOL Conference 2010 Boston Microsoft Technical Computing The Advancement of Parallelism Tom Quinn, Technical Computing Partner Manager 21 1.2 x 10 New Bytes of Information in 2010 Source:
David Chappell Data in a PaaS World A Guide for New Applications Sponsored by Microsoft Corporation Copyright 2016 Chappell & Associates Contents The Rise of PaaS Data Services... 3 The Value of PaaS for
supercomputing. simplified. INTRODUCING WINDOWS HPC SERVER 2008 R2 SUITE Windows HPC Server 2008 R2, Microsoft s third-generation HPC solution, provides a comprehensive and costeffective solution for harnessing
Course 20532A: Developing Microsoft Azure Solutions Page 1 of 7 Developing Microsoft Azure Solutions Course 20532A: 4 days; Instructor-Led Introduction This course is intended for students who have experience
Cloud Computing em Azure e Branco Luis Alves Martins Architect Advisor Microsoft email@example.com cloud computing hype? cloud computing is about where applications live on premises I purchase
An HPC Application Deployment Model on Azure Cloud for SMEs Fan Ding CLOSER 2013, Aachen, Germany, May 9th,2013 Rechen- und Kommunikationszentrum (RZ) Agenda Motivation Windows Azure Relevant Technology
Windows Azure and private cloud Joe Chou Senior Program Manager China Cloud Innovation Center Customer Advisory Team Microsoft Asia-Pacific Research and Development Group 1 Agenda Cloud Computing Fundamentals
Cost Savings Solutions for Year 5 True Ups US Dept. of Energy EA Affigent/CDWG/Microsoft Realizing Cost Savings Now and Moving to a Dynamic Datacenter via your Current EA Enterprise Desktop Solutions to
UT DALLAS Erik Jonsson School of Engineering & Computer Science Cloud Computing Trends What is cloud computing? Cloud computing refers to the apps and services delivered over the internet. Software delivered
Taking the cloud to your datacenter Microsoft Azure Stack Version 1.0 1/29/2016 CONTENTS Cloud is a paradigm, not a place... 2 Cloud computing on your terms... 3 Microsoft Azure Stack vision... 4 Reinventing
Lincoln Land Community College Capital City Training Center 130 West Mason Springfield, IL 62702 217-782-7436 www.llcc.edu/cctc Developing Microsoft Azure Solutions 20532A; 5 days Course Description This
TOOLS FOR TEAM DEVELOPMENT: WHY VENDORS ARE FINALLY GETTING IT RIGHT DAVID CHAPPELL DECEMBER 2008 SPONSORED BY MICROSOFT CORPORATION COPYRIGHT 2008 CHAPPELL & ASSOCIATES Most software development is done
W H I T E P A P E R Increasing Your VDI Project s Return on Investment Using Workspace Virtualization W H I T E P A P E R Increasing Your VDI Project s Return on Investment Using Workspace Virtualization
WHITE PAPER Meeting the Needs of Database Management for SaaS: Oracle Database 12c Sponsored by: Oracle Corp. Carl W. Olofson September 2014 IDC OPINION The move of ISV applications to the cloud is growing
Course 20532B: Developing Microsoft Solutions Five Days, Instructor-Led About this Course This course is intended for students who have experience building vertically scaled applications. Students should
Platforms in the Cloud Where Will Your Next Application Run? Jazoon, Zurich June 2011 Copyright 2011 Chappell & Associates An Organization without Cloud Computing Users A A A VM VM VM A A A Application
Description Course Summary This course provides students with the knowledge and skills to manage and deploy Microsoft HPC Server 2008 clusters. Objectives At the end of this course, students will be Plan
Learning Cloud Computing: What IT Professionals Need to Know Cloud computing promises new career opportunities for IT professionals. In many cases, existing core skill sets transfer directly to cloud technologies.
Course M20532 5 Day(s) 30:00 Hours Developing Microsoft Azure Solutions Audience The candidates targeted by this training have basic experience in implementing and monitoring Microsoft Azure solutions.
Microsoft Dynamics NAV 2013 R2 Sizing Guidelines for Multitenant Deployments February 2014 Contents Microsoft Dynamics NAV 2013 R2 3 Test deployment configurations 3 Test results 5 Microsoft Dynamics NAV
Systems Integrators in the Cloud Era Embracing the Future @DChappellAssoc Copyright 2014 Chappell & Associates SIs in the Cloud Era The SI world is changing The cloud brings new opportunities and new threats
David Chappell Understanding NoSQL Technologies on Windows Azure Sponsored by Microsoft Corporation Copyright 2013 Chappell & Associates Contents Data on Windows Azure: The Big Picture... 3 Windows Azure
System Models for Distributed and Cloud Computing Dr. Sanjay P. Ahuja, Ph.D. 2010-14 FIS Distinguished Professor of Computer Science School of Computing, UNF Classification of Distributed Computing Systems
INTRODUCTIONS Storm Technology Who we are and what we do David Chappell IT strategist and international advisor The Cloud A Rational Perspective The cloud platforms An objective overview of the Windows
Cloud Building Blocks: Paving the Road By understanding what provides a strong foundation for the cloud, CIOs can help make their deployments smooth and cost-effective. Companies are turning to cloud computing
Open Source Technologies on Microsoft Azure A Survey @DChappellAssoc Copyright 2014 Chappell & Associates The Main Idea i Open source technologies are a fundamental part of Microsoft Azure The Big Questions
IBM Global Technology Services September 2007 NAS systems scale out to meet Page 2 Contents 2 Introduction 2 Understanding the traditional NAS role 3 Gaining NAS benefits 4 NAS shortcomings in enterprise
International Journal of Information and Computation Technology. ISSN 0974-2239 Volume 4, Number 2 (2014), pp. 139-144 International Research Publications House http://www. irphouse.com /ijict.htm Cloud
How to Ingest Data into Google BigQuery using Talend for Big Data A Technical Solution Paper from Saama Technologies, Inc. July 30, 2013 Table of Contents Intended Audience What you will Learn Background
Your Guide to VMware Lab Manager Replacement white paper BROUGHT TO YOU BY SKYTAP 2 Your Guide to VMware Lab Manager Replacement Contents Your Guide to VMware Lab Manager Replacement... 3 The Power and
Planning the Migration of Enterprise Applications to the Cloud A Guide to Your Migration Options: Private and Public Clouds, Application Evaluation Criteria, and Application Migration Best Practices Introduction
CLOUD COMPUTING & WINDOWS AZURE WORKSHOP Overview This workshop is an introduction to cloud computing and specifically Microsoft s public cloud offering in Windows Azure. Windows Azure has been described
David Chappell APPLICATION PLATFORMS AND BUSINESS STRATEGY MAKING THE CONNECTION Sponsored by Microsoft Corporation Every organization has goals it s trying to reach. A business might wish to improve its
Technical Overview of Windows HPC Server 2008 Published: June, 2008, Revised September 2008 Abstract Windows HPC Server 2008 brings the power, performance, and scale of high performance computing (HPC)
Windows Server 2008 R2 Hyper-V Live Migration Table of Contents Overview of Windows Server 2008 R2 Hyper-V Features... 3 Dynamic VM storage... 3 Enhanced Processor Support... 3 Enhanced Networking Support...
The Definitive Guide to the Cloud and Kentico CMS THOMAS ROBBINS Contents Introduction... 4 What is Cloud Computing?... 4 The Benefits of the Cloud... 6 Full Hardware Utilization... 6 Lower Power Costs...
David Chappell INTRODUCING AZURE MACHINE LEARNING A GUIDE FOR TECHNICAL PROFESSIONALS Sponsored by Microsoft Corporation Copyright 2015 Chappell & Associates Contents What is Machine Learning?... 3 The
Copyright 2013 Microsoft Corporation. All rights reserved. Except where otherwise noted, these materials are licensed under the terms of the Apache License, Version 2.0. You may use it according to the
WHITE PAPER Purchase-to-pay by ReadSoft. Automated account payables processing software Author: Wille Dahl March 2012 Copyright ReadSoft - www.readsoft.com Processing invoices in the cloud or on premises
..................................... PEPPERDATA IN MULTI-TENANT ENVIRONMENTS technical whitepaper June 2015 SUMMARY OF WHAT S WRITTEN IN THIS DOCUMENT If you are short on time and don t want to read the
Copyright 2014 Microsoft Corporation. All rights reserved. Except where otherwise noted, these materials are licensed under the terms of the Apache License, Version 2.0. You may use it according to the
Creative Shorts: Twelve lifecycle management principles for world-class cloud development Foundations for optimal development on and for the cloud A Creative Intellect Consulting Shorts Report Series (ALM)
Private Cloud Computing Essentials The 2X Private Cloud Computing Essentials This white paper contains a brief guide to Private Cloud Computing. Contents Introduction.... 3 About Private Cloud Computing....
Cloud Computing Today What You Need to Know Now Copyright 2013 Chappell & ssociates n Organization without Cloud Computing Users Compute/Storage/Network pplication On-Premises Data Center Virtual Machine