More about. Front cover. ibm.com/redbooks. Prepare your WebSphere Web site for e-business on demand

Transcription

1 Front cover More about High-Volume Web Sites Prepare your WebSphere Web site for e-business on demand Architecture for virtualization with WebSphere Application Server, V5 Resilience of WebSphere Portal clusters under load Authored by the High-Volume Web Sites Team ibm.com/redbooks

2

3 International Technical Support Organization More about High-Volume Web Sites March 2004 SG

4 Note: Before using this information and the product it supports, read the information in. Second Edition (March 2004) This edition includes papers published by IBM s High-Volume Web Sites team. Copyright International Business Machines Corporation All rights reserved. Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp.

5 Contents Notices vii Trademarks viii Preface ix About this redbook xi The team that wrote this redbook xii Become a published author xiii Comments welcome xiii Chapter 1. Prepare your WebSphere Web site for e-business on demand What are leading-edge customers doing? IBM Server Allocation for WebSphere Application Server An illustration of the server allocation process Start now to be ready Introduction to grid computing and autonomic computing Grid computing and Web services standards Autonomic computing References Chapter 2. Architecture for virtualization with WebSphere Application Server, V Introduction Application server virtualization Virtualization in the application server environment Challenges Preparing the application server environment for virtualization WebSphere Application Server, Version 5.0 and virtualization WebSphere Application Server Network Deployment Clustering Workload management Administrative model Coexistence Architecture for virtualization Application isolation Sharing versus isolation -- the trade-offs System administration Sample architecture Other virtualization techniques LPARs in a virtualized server environment Conclusion Sample script References Chapter 3. Advanced clustering techniques for maximizing Web site availability with WebSphere Application Server, Version Introduction The Web server tier Clustering the Web server tier Maintenance The application server tier Copyright IBM Corp All rights reserved. iii

6 Clustering the application server tier Maintenance and failover scenarios Summary of best practices Conclusion and future directions Sample scripts References Chapter 4. Resilience of WebSphere Portal clusters under load Introduction Workload management Failover and high availability Workload management using the IBM HTTP Server version 1.x on AIX Test configuration Hardware Software Workload management in the test scripts Test descriptions Test methodology Failover test Failover test Failover test Workload management observations Examples of workload management Examples of round robin load balancing Recent example of round robin load balancing Minimizing the user impact caused by an unresponsive clone Best practices advice Capacity planning System tuning Software levels WebSphere Application Server plug-in parameters WebSphere Application Server parameters WebSphere Portal parameters Other hints and tips Summary of recommended parameters Conclusions Software configuration Technique for logging plug-in decision-making Problem overview Solution overview References Chapter 5. How WebSphere caches dynamic content for high-volume Web sites Introduction Caching dynamic content What should be cached? Where should caching take place? How is cache invalidated? WebSphere Application Server dynamic cache service Servlet/JSP Result Cache Command Cache Replication support Invalidation support iv More about High-Volume Web Sites

7 Edge of Network Caching support Tools Conclusion References Chapter 6. Impact of object serialization and local Enterprise JavaBeans on application server performance Introduction Serialization in J2EE applications Uses of serialization Using Java serialization Serialization algorithm Serialization issues Related research on improving serialization performance Test case implementation Application Environment Results Analysis Performance considerations for application design Remote interface designs HTTP session objects Reducing serialization costs Understanding workloads Conclusions Listing of the objects used with their sizes and object hierarchy References Chapter 7. Using IBM s Content Manager to manage Web content Introduction to IBM s Content Manager Content Manager components Advantages of using Content Manager for Web content management (WCM) Building a WCM application using Content Manager Author time application development Publishing application Personalization Conclusion References Chapter 8. Building a custom Web content management solution with IBM Content Manager for Multiplatforms, Version Introduction Why IBM Content Manager for Multiplatforms, Version Implementing content management at a high-volume Web site Requirements Design decisions Content Manager Lessons learned and best practices General query guidelines Query optimization Object access Transactions Connection pooling parameters Datastore pooling parameters Contents v

8 Conclusion Chapter 9. High-Volume Web Site Performance Simulator for WebSphere Introduction Using the HVWS Simulator Examples of how the HVWS Simulator is used Simulator input panels and sample output Select workload pattern Specify performance objectives Specify the hardware used or projected for use Specify the software components used or projected to be used Calculate results Graph results Display pie chart References Chapter 10. Sametime Links Scalability Report Sametime Links architecture Sample configurations Test script Hardware configuration Software configuration Services that were not tested Results Conclusions Results / raw data and graphs Test I Test II Test III Test IV Test V Test VI References Related publications IBM Redbooks Online resources How to get IBM Redbooks Index vi More about High-Volume Web Sites

9 Notices This information was developed for products and services offered in the U.S.A. IBM may not offer the products, services, or features discussed in this document in other countries. Consult your local IBM representative for information on the products and services currently available in your area. Any reference to an IBM product, program, or service is not intended to state or imply that only that IBM product, program, or service may be used. Any functionally equivalent product, program, or service that does not infringe any IBM intellectual property right may be used instead. However, it is the user's responsibility to evaluate and verify the operation of any non-ibm product, program, or service. IBM may have patents or pending patent applications covering subject matter described in this document. The furnishing of this document does not give you any license to these patents. You can send license inquiries, in writing, to: IBM Director of Licensing, IBM Corporation, North Castle Drive Armonk, NY U.S.A. The following paragraph does not apply to the United Kingdom or any other country where such provisions are inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS PUBLICATION "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF NON-INFRINGEMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Some states do not allow disclaimer of express or implied warranties in certain transactions, therefore, this statement may not apply to you. This information could include technical inaccuracies or typographical errors. Changes are periodically made to the information herein; these changes will be incorporated in new editions of the publication. IBM may make improvements and/or changes in the product(s) and/or the program(s) described in this publication at any time without notice. Any references in this information to non-ibm Web sites are provided for convenience only and do not in any manner serve as an endorsement of those Web sites. The materials at those Web sites are not part of the materials for this IBM product and use of those Web sites is at your own risk. IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring any obligation to you. Information concerning non-ibm products was obtained from the suppliers of those products, their published announcements or other publicly available sources. IBM has not tested those products and cannot confirm the accuracy of performance, compatibility or any other claims related to non-ibm products. Questions on the capabilities of non-ibm products should be addressed to the suppliers of those products. This information contains examples of data and reports used in daily business operations. To illustrate them as completely as possible, the examples include the names of individuals, companies, brands, and products. All of these names are fictitious and any similarity to the names and addresses used by an actual business enterprise is entirely coincidental. COPYRIGHT LICENSE: This information contains sample application programs in source language, which illustrates programming techniques on various operating platforms. You may copy, modify, and distribute these sample programs in any form without payment to IBM, for the purposes of developing, using, marketing or distributing application programs conforming to the application programming interface for the operating platform for which the sample programs are written. These examples have not been thoroughly tested under all conditions. IBM, therefore, cannot guarantee or imply reliability, serviceability, or function of these programs. You may copy, modify, and distribute these sample programs in any form without payment to IBM for the purposes of developing, using, marketing, or distributing application programs conforming to IBM's application programming interfaces. Copyright IBM Corp All rights reserved. vii

10 Trademarks The following terms are trademarks of the International Business Machines Corporation in the United States, other countries, or both: AIX DB2 Domino e-business on demand IBM iseries Lotus MQSeries pseries Redbooks(logo) RS/6000 Sametime Tivoli VisualAge WebSphere xseries zseries The following terms are trademarks of other companies: LoadRunner is a registered trademark of Mercury Interactive. Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the United States, other countries, or both. Java and all Java-based trademarks and logos are trademarks or registered trademarks of Sun Microsystems, Inc. in the United States, other countries, or both. UNIX is a registered trademark of The Open Group in the United States and other countries. Other company, product, and service names may be trademarks or service marks of others. viii More about High-Volume Web Sites

11 Preface In 1999, Dr Willy Chiu, noticing the emergence of many large and complex customer Web sites, and observing how they often failed to deliver the expected robustness and customer satisfaction, set up a new team within IBM called the High-Volume Web Sites (HVWS) team. This team was chartered to work with customers and IBM internal teams involved in the creation and deployment of really large Web sites, and in particular those supporting e-business applications used by customers over the Internet -- typical business to customer (B2C) applications but large and complex ones. The team would learn from this experience and document proven best practices so that customers could learn how to make high-volume Web sites that worked well, and so that IBM developers could improve their products to better support high- volume sites. On December 9, 2003 IBM announced that it has chosen Silicon Valley as the location for its first "e-business on demand Center of Competency" -- which will amass heavy technology resources and expertise to help companies advance their Internet initiatives. The Center of Competency in IBM's Silicon Valley Lab will be the first of several to open in the next year as part of IBM Chairman and Chief Executive Officer Samuel J. Palmisano's $10 billion commitment for research, acquisitions, marketing and training centers devoted to e-business on demand. This is IBM's one-year-old initiative to help companies improve efficiency, productivity, and their ability to respond to changing conditions such as peaks in customer demand. IBM is beefing up capabilities in the Silicon Valley lab to help companies that operate some of the world's busiest Web sites and must reliably handle huge, often unpredictable amounts of traffic (with page views that can soar into the hundreds of millions each day). The center will house experts from IBM Software, IBM Research, and other parts of the company who are some of the world's top experts in high-performance Internet computing, as well as hundreds of server computers and other equipment that will allow companies to design and test new technologies for on-demand computing. The center is an extension of IBM's existing High-Volume Web Site Lab in the Silicon Valley Lab, which has worked with ebay, Charles Schwab, CIGNA, Federated Department Stores, VISA and many others to stretch the boundaries of business computing on the Internet. For more than four years, IBM's High-Volume Web Sites (HVWS) team has been working with many of the world's largest Web sites. The team has accumulated a significant amount of knowledge and defined best practices for designing and deploying high-volume sites, earning a reputation as one of the world's leading centers of expertise on scalable e-business infrastructures. The team has locations in California, New York, Japan, Korea, China, Taiwan, and the United Kingdom. The IT infrastructures that comprise most high-volume sites present unique challenges in design, implementation, and management. While actual implementations vary, Figure 0-1 shows a typical e-business infrastructure comprised of several tiers. Each tier handles a particular set of functions, such as serving content (Web servers, such as the IBM HTTP Server), providing integration business logic (Web application servers, such as the WebSphere Application Server), or processing database transactions (transaction and database servers). Site workloads are assumed to be high volume, serving dynamic, volatile data. Copyright IBM Corp All rights reserved. ix

12 Users Edge server Web server Application server Database server & legacy systems Internet Figure 0-1 Multi-tier infrastructure for e-business The HVWS team analyzes site traffic patterns to improve performance and availability. Figure 0-2 shows how IBM's HVWS team defines the life cycle of a Web site; it shows also the categories of best practices recommended for one or more phases of the cycle. Life cycle of a Web site Planning Architecture Design Build & Test Deployment Service delivery Preparing for e-business on demand Design for scalability Plan for growth Design pages for performance Manage end-to-end performance Design for availability Figure 0-2 Life cycle of a Web site As it accumulates experience and knowledge, the HVWS team publishes papers aimed at helping CIOs and others like you understand and meet the new challenges presented during one or more of the phases. This IBM Redbook is a compilation of the HVWS papers, which are available individually at the HVWS Web page. In January 2004, IBM renamed the HVWS team to reflect its success and expanding mission. The new name is High Performance On Demand Solutions (HiPODS) in recognition of the evolution of its work with some of IBM's largest customers, including ebay and Charles Schwab, whose projects are evolving from high volume Web architectures toward high performance on demand operating environments, including all the software brands. x More about High-Volume Web Sites

13 About this redbook This redbook is a compilation of the HVWS white papers published in 2003: Chapter 1, Prepare your WebSphere Web site for e-business on demand, introduces how some of IBM's largest customers are starting their transition to on demand computing. It also introduces a limited IBM offering, IBM Server Allocation for WebSphere Application Server, developed by the HVWS group. Finally, the chapter identifies what you can do now to improve resource use and ready your e-business to benefit from on demand computing. Chapter 2, Architecture for virtualization with WebSphere Application Server, Version 5, explains how enterprises can use the new and enhanced features of WebSphere Application Server, Version 5 as the first steps in building and realizing the value of resource virtualization today. Enterprises who virtualize will be ready to implement emerging on-demand offerings such as IBM Server Allocation for WebSphere Application Server. Chapter 3, Advanced clustering techniques for maximizing Web site availability with WebSphere Application Server, Version 5, discusses advanced techniques for Web and application server clustering using IBM WebSphere Application Server, Version 5. Server clustering is critical to the on-demand operating environment and, in particular, to the Web infrastructure. Server clustering can be used to help achieve continuous availability of Web sites in the always-on global marketplace. Chapter 4, Resilience of Websphere Portal clusters under load, describes a HVWS project to test and validate the resilience of WebSphere Portal clusters under realistic load conditions against a variety of failure scenarios. It includes a description of the behavior of the clusters during recovery of application server clones and nodes, verifying the successful redistribution of workload to the recovered clones and/or nodes. This chapter includes lessons learned and a number of best practices settings and guidelines. Chapter 5, How WebSphere caches dynamic content for high-volume Web sites, introduces a feature of WebSphere Application Server Version 5 that can improve Web Site performance. Optimizing for scalability and personalization remains a significant challenge for e-businesses as they balance the demands for instantaneous changes, high performance, availability, reliability, and security. Vendors are responding with infrastructure options and supporting hardware and software platforms that address these requirements. Chapter 6, Impact of object serialization and local enterprise JavaBeans on application server performance, explains the Java serialization process, related research, and the cost of serialization and deserialization of Java objects within WebSphere Application Server, Version 5.0 environments. The objective of this chapter is to help software architects and designers make informed choices when implementing remote objects. Chapter 7, Using IBM s Content Manager to manage Web content, describes a joint project between IBM and a customer to use IBM Content Manager to develop in a short time a customized application to manage Web content. Leveraging the capabilities of IBM Content Manager avoids issues common to the development of applications that manage Web content. Chapter 8, Building a custom Web content management solution with IBM Content Manager for Multiplatforms, Version 8, describes a joint project between IBM and a customer to use IBM Content Manager to develop in a short time a customized application to manage Web content. Leveraging the capabilities of IBM Content Manager avoids issues common to the development of applications that manage Web content. Chapter 9, High-volume Web site Performance Simulator for WebSphere, introduces the High-Volume Web Site Performance Simulator for WebSphere, an analytic queuing model Preface xi

14 that estimates the performance of a Web server based on workload patterns, performance objectives, and specified hardware and software. The results can be used as guidelines for configuration sizing. Chapter 10, Sametime Links scalability report, reviews the results of a test of the scalability of Sametime Links. The objective of the test was to understand how Sametime Links could be scaled to support a very large number of users. Overall, the tests showed that Sametime could support 300,000 concurrent Sametime Links clients in the test environment. The team that wrote this redbook Many IBMers contributed effort and information to these chapters and projects. Table 0-1 lists the primary technical author(s) for each chapter. For a complete list of contributors, see the corresponding white paper at Table 0-1 Primary authors. Chapter Title Primary author(s) 1 Prepare your WebSphere Web site for e-business on demand 2 Architecture for virtualization with WebSphere Application Server, Version 5 3 Advanced clustering techniques for maximizing Web site availability with WebSphere Application Server, Version 5 Luis Ostdiek Luis Ostdiek, Catherine Diep Luis Ostdiek 4 Resilience of WebSphere Portal clusters under load Alastair Watson, Paul Edlund 5 How WebSphere caches dynamic content for high-volume Web sites 6 Impact of object serialization and local enterprise JavaBeans on application server performance Catherine Diep Ranjit Nayak 7 Using IBM s Content Manager to manage Web content Rekha Garpati 8 Building a custom Web content management solution with IBM Manager for Multiplatforms, Version 8 9 High-Volume Web Site Performance Simulator for WebSphere Chin Huang Noshir Wadia 10 Sametime Links scalability report Harold Hall, Amir Perman Thanks to the following people for their contributions to this Redbook project: Susan Holic, IBM HVWS Team, Silicon Valley Laboratory, San Jose, California Linda Legregni, Program Director, HVWS Team Project Office, Silicon Valley Laboratory, San Jose, California Joe DeCarlo, Manager, International Technical Support organization, San Jose, California xii More about High-Volume Web Sites

15 Become a published author Join us for a two- to six-week residency program! Help write an IBM Redbook dealing with specific products or solutions, while getting hands-on experience with leading-edge technologies. You'll team with IBM technical professionals, Business Partners and/or customers. Your efforts will help increase product acceptance and customer satisfaction. As a bonus, you'll develop a network of contacts in IBM development labs, and increase your productivity and marketability. Find out more about the residency program, browse the residency index, and apply online at: ibm.com/redbooks/residencies.html Comments welcome Your comments are important to us! We want our Redbooks to be as helpful as possible. Send us your comments about this or other Redbooks in one of the following ways: Use the online Contact us review redbook form found at: ibm.com/redbooks Send your comments in an Internet note to: redbook@us.ibm.com Mail your comments to: IBM Corporation, International Technical Support Organization Dept. 1WLB Building 662 P.O. Box Research Triangle Park, NC Preface xiii

16 xiv More about High-Volume Web Sites

17 1 Chapter 1. Prepare your WebSphere Web site for e-business on demand Executive summary: In the last few years, IBM's High-Volume Web Site (HVWS) team has engaged with many large customers to help them understand their workloads, and to design and implement highly scalable e-business infrastructures. The resulting systems operate now to meet the unpredictable workloads common to an e-business. When not meeting peak demands, however, these resulting infrastructures are often left with excess capacity. Excess capacity and the increasing pressure to improve the return on investment are combining to drive businesses to seek new opportunities to use existing resources. Several leading-edge customers are already working with IBM to optimize their resource use and to consider how emerging on demand technologies and standards can be part of the solution. While the technologies and standards are still under development, there are actions to take now so that when the technologies and standards are available, you'll be ready to transform to on demand computing and benefit from its advantages of managing workloads dynamically, adding applications without adding computing resources, and reducing human intervention. Among the key technologies for enabling e-business on demand are grid computing and autonomic computing. A grid is a collection of distributed computing resources available over a network that appears to an end user or application as one large virtual computing system. A grid can span locations, organizations, machine architectures, and software boundaries offering the promise of providing virtually unlimited power, collaboration, and information access to everyone connected to the grid. One effect of grid computing can be to make network computing more like a utility. You deliver computing power to where you need it only when you need it; you pay for what you use, when you use it. In addition to grid computing, autonomic computing is moving from concept to reality using Web services and Java 2 Enterprise Edition (J2EE) programming constructs. Autonomic computing is an approach to self-managed computing systems with minimal human intervention. Using a combination of grid and autonomic computing, infrastructures will be able to monitor current workload, analyze it versus historical trends and available resources, plan the reallocation of those resources, and automatically execute the movement of workload or resources to maximize computing responsiveness to meet service level Copyright IBM Corp All rights reserved. 1

18 objectives; in autonomic terms, this is called the autonomic control loop. This will allow server farms to act as a single virtual computing resource. Having aligned itself around e-business on demand, IBM has new offerings and roadmaps for virtualization technologies that encompass servers, storage, networks, and distributed applications. IBM makes it possible for enterprises to virtualize WebSphere resources today, so that they don't have to wait until tomorrow to realize the business value of the on-demand operating environment. See Chapter 2 to learn how enterprises can use the new and enhanced WebSphere features as the first steps in building and realizing the value of resource virtualization. Enterprises that virtualize will be ready to implement emerging on-demand offerings such as IBM Server Allocation for WebSphere Application Server, a limited offering developed by IBM Research and the HVWS group. This chapter introduces the early work of some of our customers and summarizes the IBM Server Allocation for WebSphere Application Server. The chapter identifies what you can do now to improve resource use and ready your e-business to benefit from on demand computing. The last section provides a quick introduction to grid computing and autonomic computing and identifies where you can learn more. What are leading-edge customers doing? IBM's largest customers are finding ways to use their excess capacity as a part of their transition to e-business on demand. With help from IBM, they determine which application is the best candidate for their early work. Here are examples of three customers who have begun their transition. A leading financial company that offers online trading wants to be able to 'borrow' resources from the high-priority trading application while maintaining its service level agreements (SLA) for trading. The company plans to provide online self-service advice using the excess, available resource. The advice application will run as a lower priority application and be as responsive as it can be with available resources, which will vary based on the needs of the trading application. This application of grid and autonomic computing enables the company to offer their customers more with no additional investment, thereby improving the company's return on investment and improving its ability to compete. A large insurance company has the objective of making all servers available to all applications, essentially making their server farm a single virtual computing resource. By doing so, the excess capacity will serve as a 'shock absorber' during peaks in workload. Server grids facilitate workload management by eliminating the need to plan for and manage standby capacity. An automobile manufacturer has the vision of allowing a potential customer to design a car online. The new technologies enable such visions without investing in additional capacity. The limited IBM offering called IBM Server Allocation for WebSphere Application Server addresses these needs using open Web services standards, such as Web Services Description Language (WSDL) and Simple Object Access Protocol (SOAP). The next section introduces the technology. 2 More about High-Volume Web Sites

19 IBM Server Allocation for WebSphere Application Server The HVWS team and IBM Research are developing on demand technology using constructs of the autonomic control loop and open Web services standards. IBM Server Allocation for WebSphere Application Server enables WebSphere customers to balance workloads and allocate resources on demand. IBM intends to enhance it as the emerging Open Grid Standard Architecture (OGSA) protocols mature, and make it an integrated component of a future version of the WebSphere Application Server. IBM Server Allocation allows the customer to run multiple transactional applications and one parallel application in the same grid infrastructure. As the traffic of a higher priority workload increases to such a level that the service level agreement (SLA) objective cannot be met, the system drains the lower priority workloads to other servers in the local or remote grid, to make additional servers available for high priority transactions. When all the servers in the WebSphere server pool are at a high use level serving high priority applications, IBM Server Allocation for WebSphere integrates with IBM Tivoli Intelligent ThinkDynamic Orchestrator to provision additional servers from the data center server pool. As the traffic of the higher priority workloads decreases, the system schedules the lower priority workloads on to the servers where they were previously running. The primary components of the offering, as shown in Figure 1-1, are defined to support the goals of the autonomic control loop. IBM Server Allocation for WebSphere Server Allocation Manager WSLA Monitor M A SLA monitor & analysis for each workload P Policy Planner Planning for execution E Execution Engine Capacity allocation/deallocation in the WebSphere server pool Server Allocation Resource Manager Resource Parallel application Manager scheduler (PAS) TAGSS Workload Provisioning scheduling new servers into the WebSphere server pool Application using IBM Provisioner Tivoli ThinkDynamic Allocate/deallocate WAS Intelligent Orchestrator Server Allocation Setup Manager Documentation GUI for configuration and SLA definition Autonomic control loop Analyze Plan Monitor Execute Knowledge Key: M A P E Monitor Analyze Plan Figure 1-1 Components of the IBM Server Allocation for WebSphere IBM Server Allocation has three major components: The server allocation manager component is the most sophisticated of the three. It provides the autonomic control loop functions. It contains subcomponents that monitor Chapter 1. Prepare your WebSphere Web site for e-business on demand 3

20 workloads, analyze data to determine if the workload will continue to meet its service level agreement, and if needed, plans whether resources need to be reallocated, and invokes the server allocation resource manager to reallocate as needed. The server allocation resource manager allocates resources as needed to meet service level objectives. Its major subcomponent, the parallel application scheduler (PAS), decomposes processing of parallel applications into tasks that can be executed on the available CPUs in parallel, distributes these tasks across the CPUs, and gathers and combines the results. When all the servers in the WebSphere server pool are at capacity, the server allocation resource manager calls IBM Tivoli Intelligent ThinkDynamic Orchestrator to provision a server from the data center server pool into the WebSphere server pool. The server allocation setup manager facilitates setting service level agreements, and controlling the installation and operation of the Server Allocation components. It is fully integrated with the WebSphere Application Server administration framework. Server Allocation includes WebSphere Studio Application Developer plug-ins to help in the development of new parallel applications or allow you to modify existing parallel programs that are written in C, Fortran, or other programming languages. Customers can also easily develop Java applications with a new API, so that the application can take advantage of the servers in the WebSphere environment. An illustration of the server allocation process This section illustrates the server allocation process. The scenario is an infrastructure for an online brokerage providing two transaction workloads and one parallel workload: One transactional workload simulates the common operations of online stock trading. This class of transactional workload represents business critical applications. The second transaction workload is an account management application through which brokerage customers maintain and view their account information. This application has a lower priority than the stock trading application, but is of higher priority than the parallel application. The parallel workload simulates financial analysis services that commercial brokerage houses offer to their customers, such as retirement portfolio analysis. Parallel applications can be broken into units that can be processed in parallel on multiple machines; their processing is typically complex and numerically intensive. The more CPUs made available to a parallel application, the faster the application runs. This class of applications falls in the value-add category, contrasted with those in the business critical category. Our illustration assumes a business rule that processing transactional workloads has higher priority than processing parallel workloads. This business rule implies that whenever there is a surge in traffic to the business critical application, resources are taken away from the value-add application and assigned to the business critical application. Figure 1-2 presents a high-level overview of an infrastructure using server allocation. 4 More about High-Volume Web Sites

21 Multiple Web transactions Real time High priority Critical SLA Parallel applications Complex Numerically NE Feature! IBM Server Allocation for WebSphere Application Server IBM Tivoli Intelligent ThinkDynamic Orchestrator WebSphere server pool Data center server pool Figure 1-2 High-level overview of an infrastructure using server allocation Figures 1-3 through 1-7 illustrate the server allocation process and how activity might be displayed as it happens. Figure 1-3 introduces what we call the dashboard. The dashboard indicates how server allocation proceeds through the process of monitoring server use, determining through analysis which servers can no longer meet the SLA requirement, reducing the number of servers assigned to the parallel application, reallocating those servers to the WebSphere Network Dispatcher for use by high priority applications until the load returns to SLA limits, and returning the reallocated servers to the parallel application. When indicated, server allocation can provision an application to a server in the pool or, if necessary, invoke the Tivoli Intelligent ThinkDynamic Orchestrator to provision an additional server from the data center server pool. The window labeled SAM Server Pool Monitor lists the five servers that comprise the infrastructure. For each server, the window shows what workload(s) is currently assigned, the workload priority, what applications are installed, and current CPU use. Note that Servers 14 and 15 are dedicated to the higher priority workloads, while Server 16, 17, and 18 can run more than one workload. The assignments and metrics are updated periodically as the system responds to traffic and manages the service level requirements for each workload by dynamically allocating and reallocating the servers. Chapter 1. Prepare your WebSphere Web site for e-business on demand 5

22 Figure 1-3 Server allocation dashboard; servers are well below max SLAs in light traffic During light traffic: Server 14 is assigned to the account management workload (blue), the priority 2 application Server 15 is assigned to the stock trading workload (green), the priority 1 application Servers 16, 17, and 18 are assigned to the portfolio workload (purple), the priority 3 application The Transactional Workload Throughput window shows how the server pool is processing the traffic for each workload. The throughput scale changes as the throughput changes. In the same color as the workload designation above, the horizontal lines represent the minimum and maximum service level for the workload. The vertical bars, again in the same workload color, indicate average throughput (dark bar) and total throughput (light bar). On the right of the dashboard, there are two driver windows, one for each transactional application. The drivers are used to simulate the workload traffic; traffic is low, around 40 concurrent users per application. Figure 1-4 illustrates what happens when the traffic increases. For our illustration, we simulate 300 concurrent users for the Stock Trading workload and 200 for the Account Manager workload. As traffic increases, the available resources are used to such an extent that the service level agreements for both Stock Trading and Account Manager are exceeded. To handle the increased traffic and maintain the SLA, Server Allocation has already assigned 6 More about High-Volume Web Sites

23 Servers 16 and 17 to Stock Trading, making those additional resources available to the high priority application. Account Manager is also missing its SLA Figure 1-4 Max SLAs are exceeded; Server Allocation will provision an application to Server 18 However, the traffic is such in this snapshot that even the combined resources of Servers 15, 16, and 17 can no longer maintain the SLA. The average Stock Trading throughput still exceeds the maximum SLA threshold. Look at the line for Server 18. The (P) Trade indicator in the Workloads Installed column indicates that Server Allocation is provisioning Server 18 with the Stock Trading application to provide the additional resource to process Stock Trading traffic. Chapter 1. Prepare your WebSphere Web site for e-business on demand 7

24 The snapshot in Figure 1-5 now shows that Servers 15, 16, 17, and 18 are dedicated to Stock Trading. The bar graph for the last timing interval also shows that the SLA is still not being met, even after taking over the last server in the WebSphere server pool. Account Manager continues to miss its SLA. When this situation occurs, Server Allocation invokes the Tivoli Intelligent ThinkDynamic Orchestrator, which provisions another server from the data center pool. Figure 1-5 All installed servers are at capacity; an additional server is needed to maintain SLA 8 More about High-Volume Web Sites

25 The snapshot in Figure1-6 shows Server 19 has been added to the list of servers supporting the online brokerage. Most importantly, the combined resources of the servers allocated to Stock Trading are now consistently meeting the SLA. Figure 1-6 An additional server is processing stock trades; max SLAs are being maintained Start now to be ready It's not too early to begin your transition/transformation to an e-business on demand. You need a strategy to get you from where you are now to where you need to be to reap the benefits of an on demand e-business. Start with learning about the advanced technologies and, as you do, begin to create your strategy. Review the quick introductions to grid and autonomic computing in this chapter. Pursue the additional references to learn more. Then think about applying the elements of the autonomic control loop to your strategic planning. Monitor: The first step is to learn as much as you can about your current workloads. Implement workload monitors so you can better understand workload characteristics and patterns. Application measurements you want to monitor include CPU utilization, response time patterns, and workload peaks and valleys. Tivoli offers the complete IBM solution for managing the extended WebSphere environment. For precise viewing of performance metrics, users can start with the Tivoli Performance Viewer, a complimentary tool shipped with WebSphere Application Server. For ongoing production monitoring, Tivoli offers the following tools: IBM Tivoli Monitoring for Web Infrastructure (ITMfWI) provides best-practice monitoring of the key elements of WebSphere Application Server, and IBM Tivoli Monitoring for Transaction Performance (ITMfTP) provides a unique perspective of monitoring, that of the end-user. Chapter 1. Prepare your WebSphere Web site for e-business on demand 9