Service Management Functions
Capacity Management
On This Page
Executive SummaryAccording to the META Group, “The days of capacity/performance management (C/PM) being driven by vague corporate directions and wishful line-of-business thinking are slowly coming to an end, as besieged CIOs learn to ‘think' platform utilization and eye consolidation benefits. It is time for C/PM to seize the day and use strong processes and technical skills that can translate operational efficiencies and effectiveness into improved bottom-line performance.” Capacity management is the process of planning, analyzing, sizing, and optimizing capacity to satisfy demand in a timely manner and at a reasonable cost. This process should be proactive and responsive to business needs because the business cannot add resources after a capacity problem has happened without impacting performance. The capacity management process has a broad scope that brings together business, service, and resource capacity needs in order to ensure optimal use of the resources needed to achieve the levels of performance agreed upon with the end user or customer. “Optimal” in this context refers to resource usage at the best place, time, quantity, and price. Capacity management focuses on procedures and systems, including specification, implementation, monitoring, analysis, and tuning of information technology (IT) resources and their resulting service performance. Capacity requirements are based on qualitative and quantitative standards set by the service level management process and specified within the provisions of a service level agreement (SLA) or operating level agreement (OLA). The capacity management process relies on a set of iterative tasks—monitoring, analysis, modeling, optimizing, and change initiation—to achieve its goals. Why Is Capacity Management Important?A META Group study published in September 2003 reveals that capacity planning is the top critical issue for large enterprises (that is, those with more than 1,000 people), with 33.8 percent of respondents identifying this as a critical issue for 2003. This high priority will continue through 2005/2006, escalating consolidation and efficiency demands. Capacity management is a function of IT operations that provides guidance on how to plan, justify, and manage appropriate levels of resources needed for a given solution. Improper planning for capacity can lead to wasted resources resulting in unnecessary cost, or lack of resources resulting in poor performance or the unavailability of an IT service. Capacity management also provides input to procedures necessary for planning, implementing, and operating an efficient IT service. There is a need to monitor and interpret service capacity levels and IT component performance in order to have the capacity data essential for the planning and design of new systems, services, and business solutions. Capacity management interrelates with availability management and is key to establishing the optimal capacity for a new service or component or to changing an existing service or component. Capacity management also provides the performance and capacity planning feedback that is necessary when producing a contingency plan for business continuity. Capacity management is one of the most important quantitative aspects of an IT service. Since there is both a business and end-user focus to capacity management, once implemented this process delivers the following primary benefits:
Other benefits include:
IntroductionDocument PurposeThis guide provides detailed information about the Capacity Management service management function (SMF) for organizations that have deployed, or are considering deploying, Microsoft technologies in a data center or other type of enterprise computing environment. This is one of more than 20 SMFs defined and described in the Microsoft® Operations Framework (MOF). This guide assumes that the reader is familiar with the intent, background, and fundamental concepts of MOF as well as the Microsoft technologies described. An overview of MOF and its companion, Microsoft Solutions Framework (MSF), is available in the Overview section of the MOF Service Management Function Library document at http://www.microsoft.com/mof (click on Service Management Functions). This overview also provides abstracts of each of the SMFs defined within MOF. Detailed information about the concepts and principles of each of the frameworks is also available in technical papers the same Web site. What’s New?This document has been restructured to incorporate the latest MOF/SMF document standards. The capacity management process as described in this document has been aligned with the definitions set out in the ITIL Service Delivery book. Updated analyst research on capacity management has been included, and the relationship between the Capacity Management SMF and the new Infrastructure Engineering SMF has been incorporated. An example of capacity management is available in Appendix C. This example brings together the concepts and ideas presented in the SMF in order to help the reader understand the broad scope and value of capacity management. The appendices also include examples of a capacity plan and capacity management database (CDB). (This is related to but separate from the configuration management database, or CMDB.) Also included is a checklist for a sample capacity planning process. FeedbackPlease direct questions or comments about this SMF to msmfeed@microsoft.com. Capacity Management OverviewThe Capacity Management SMF helps organizations achieve and sustain the IT service capacity requirements they need to support their business at a justifiable cost. For the purposes of this document, the term “capacity” is one of convenience, which, depending on the context, may imply resource capacity, such as storage, processor speed, network, or human resources, or an end-to-end IT service capacity, such as messaging, customer relationship management (CRM), or order processing. Many of the principles and suggestions for best practices still apply regardless of the type of capacity being optimized. Capacity management is made up of three subprocesses:
These subprocesses all share a common set of activities that are applied from different perspectives. They include the following:
Each of these subprocesses works toward the production and maintenance of a capacity plan and triggers requests for change through the appropriate channel. These activities all support the proper management of resources and service performance levels in order to conform to current and anticipated business requirements. Goals and ObjectivesThe goal of capacity management is to optimize the capability of the IT infrastructure and supporting organization in order to deliver a cost-effective and sustained level of availability that enables the business to satisfy its objectives. For example, the right level of processor performance or storage capacity for a business-critical service needs to be carefully established in order to fulfill customer demand in a timely and cost-effective manner. The goal of capacity management is to continuously optimize existing and future IT resource demands and supply. Based on this optimizing goal, the capacity management process maintains a capacity plan and initiates requests for change that may result in environmental updates or new facilities. When new facilities are developed or purchased, they are released to operations to continue the cycle of effective operations. ScopeThe capacity management process should be the focal point for all IT performance and capacity issues. The overall responsibility for all capacity-related activity lies with a centralized or enterprise capacity management process. For both operational and developmental environments, the process should encompass the following:
Vital business operations should continue in the face of any IT disruption, whether planned or unplanned. Recovery needs to be accomplished quickly and reliably, so failures must be as transparent as possible to the customer. If the IT operation is global, then corporate IT should have a global vision and implement the appropriate, cost-justified solution. Corporate data centers, distributed servers, personal computers, and the enterprise network resources need to be considered in order to remain responsive and competitive in a global economy. It is important for the capacity management role to work with other management teams and disciplines. Capacity management provides feedback to service continuity management and works with the availability management, risk management, and financial management process owners to design new solutions when failover and recovery measures are implemented. Key DefinitionsAvailability. The amount of time an application is available to perform work, typically measured in percentage uptime. Capacity. A capability that is required for delivering the agreed-upon performance at the required service level and cost. Capacity management. Responsible for ensuring that IT processing, storage, and network capacity matches the evolving demands of the business in the most cost-effective and timely manner. Capacity management database (CDB). Contains the detailed technical, business, and service level management data that supports the capacity management process. The resource and service performance data in the database can be used for trend analysis and for forecasting and planning. Capacity plan. The consolidated output of a capacity management process. The capacity plan documents current levels of resource utilization and service performance. After considering business requirements, it forecasts future resource requirements for IT services to support them. The capacity plan recommends the resource levels and changes necessary to accomplish operating level requirements that support the service level agreement (SLA). The capacity plan includes the cost and benefit of those resources, reports of their compliance to the IT SLA, and the priority and impact of systems and resources on the overall business and the IT infrastructure. Demand management. Responsible for ensuring that the future business requirements for IT services are considered, planned, and implemented in a timely manner. The capacity management staff can achieve this by analyzing current resource utilization of the various IT solutions and generating trends and forecasts. These future requirements come from account management, which constantly probes current and future customer needs. ITIL (IT Infrastructure Library). A widely accepted approach to IT service management. ITIL provides a comprehensive and consistent set of best practices for IT service management, promoting a quality approach to achieving business effectiveness and efficiency in the use of information systems. ITIL is based on the collective experience of commercial and governmental practitioners worldwide. This has been distilled into one reliable, coherent approach, which is fast becoming a standard used by some of the world’s leading businesses. For more information on ITIL, see http://www.itil.co.uk/. Workload management. Responsible for translating customer demands into workloads required by IT solutions (the various applications used to create the actual solution) so that the required resources can be determined from this analysis. The process translates both current and future demands to workloads. Managing Business RequirementsOne of the objectives of capacity management is to deliver best practices that establish a communications framework between IT operations and the business. This is essential for ensuring that the capacity management process works toward meeting the business capacity requirements. The typical problem is that IT operations often works with measures of computer performance, megabytes of storage, or other expressions of capacity or throughput. At the same time, the business is using expressions such as sales effectiveness, market share, time to market, profit, cash flow, and return on investment (ROI). (For more information on ROI, see the Financial Management SMF guide at http://www.microsoft.com/technet/solutionaccelerators/cits/mo/smf/smffinmg.mspx.) This different terminology can lead to a misalignment between the business needs and the levels of IT service provision. In order to avoid such a misalignment, the capacity management process defines a subprocess dedicated to business capacity management. This is an ongoing process that reviews and captures business requirements and then translates them into a consistent terminology that can be used to direct service and resource capacity management activities as well as to update the capacity plan and CDB appropriately. Service level management (SLM) assists the process of managing business requirements by placing a “business value” on a particular service. The IT organization can then plan to meet the needs of that service. This in turn creates operating level requirements (OLRs) for each IT function. An OLR is an agreed-upon, measurable service metric “target” between two or more IT entities, applied to the services provided to those entities and described in an OLA. An operating level agreement (OLA) is an internal agreement between two or more IT entities. The OLA defines the responsibilities of all participating parties within IT and binds these parties to provide a particular service. The parties agree to specific levels of quality and quantity of services provided. It is similar to an SLA, but is normally not as formal. The OLA should have its metrics stored in the CDB. Defining Operating Level RequirementsCapacity management is part of the Optimizing Quadrant in the MOF Process Model. This quadrant recognizes that running IT operations successfully is a prerequisite to achieving business success and operational efficiency. The Optimizing Quadrant addresses, among other things, two specific elements of operations:
The service level management process is formalized by service level agreements (SLAs). Service level agreements must directly reflect business requirements, IT capabilities, and appropriate levels of cost. Therefore, IT must be rigorous in determining what levels of service it can reasonably agree to deliver within capability and budget. IT must then create and maintain metrics to measure the services provided. Services that cannot be measured do not meet a metric and must not be included in an SLA. (For more information, see the MOF Service Level Management SMF guide at http://www.microsoft.com/technet/solutionaccelerators/cits/mo/smf/smfslamg.mspx.) Customer expectations may exceed technical capability, so it is essential that customer expectations for new applications be managed from the outset. Capacity management provides an effective opportunity to balance customer expectations with cost and available resources to achieve appropriate and achievable target performance levels. It is useful to break down the end-to-end provision of any service into manageable pieces and to examine each of these pieces in turn. Microsoft Operations Framework breaks down the overall infrastructure that makes up and supports an IT service into the following IT categories:
This means that for each IT category identified above, operating level requirements should be generated and documented to ensure that the overall SLA is met. Some IT categories may have little, if any, impact on the delivery of a given service. Others may have a significant impact on the ability of IT to satisfy the SLA. The important thing is to thoroughly and carefully evaluate the impact that each IT category has on the overall delivery of service. Establishing Quality of Service Metrics for Capacity ManagementEffective capacity management ensures that appropriate IT resources are available when needed. With the proper management systems and tools in place for monitoring existing resource usage, thresholds can be set to alert the IT organization when conditions may indicate or lead to performance or availability issues. With the rapidity of change common in many business environments, capacity management needs to put proper tools in place to remain proactive. This includes using predictive modeling and “what-if” scenarios to estimate appropriate levels of resources and anticipating the level of performance needed to satisfy business requirements. Specific review milestones are defined and refined with the effective coordination of the service level, financial, workforce, service continuity, availability, and capacity management processes. Success in capacity management is dependent on a number of factors:
The capacity management process needs to manage its own quality and effectiveness in order to improve its operation wherever possible. Quality of Service (QoS) metrics are essential for this. Success or failure of the overall process can be measured by examining the following QoS areas. Business forecasts:
Technology:
Cost-effectiveness:
Alignment to business need:
For each of these QoS metrics, capacity management should define goals and then monitor the realization of them. The capacity manager must improve the process and manage it in such a way that the stated goals can be reached within the defined time frames. Processes and ActivitiesProcess Flow SummaryCapacity management is a continual process with several activities that are performed throughout the process. In order to keep this document brief, only a selection of the core capacity management tasks of this SMF have been chosen for a more detailed explanation (see Figure 1).  Figure 1. Capacity management as an iterative process MonitoringCapacity management involves the internal operating level requirements and associated metrics for each of the key IT layers that contribute to the overall SLA. It is important that the utilization of each resource and service be monitored on an ongoing basis to ensure that hardware and software resources are being used optimally and that all agreed-upon service levels can be achieved. Most monitoring tasks are near-term in nature and rely on underlying tools and principles for operation. The collected information must be recorded or sampled over a determined period. The amount of sampling and resources required to do so must also be examined. The capacity management database (CDB) should contain information points to identify historical trends and patterns. For more information, see the MOF Service Monitoring and Control SMF guide at http://www.microsoft.com/technet/solutionaccelerators/cits/mo/smf/smfsmc.mspx. Data must be gathered not only at the total resource utilization level, but also at a more detailed level for the workload on each particular resource. This must be carried out across the whole infrastructure, including applications, servers, and networks. Similarly, data must be collected for each service—for example, availability and user screen response time. Part of the monitoring activity is from a baseline, or profile, of the normal operating levels. If thresholds beyond the norm are exceeded, alarms are raised and exception reports are produced. These thresholds and baselines are determined from the analysis of previously recorded data and can be set on:
It is also important to remember that monitoring takes up system capacity and thus can influence the performance of the system. Focus performance measurement and monitors on client SLAs. Operating level requirements and other necessary elements for monitoring often fall out of their overall contribution to meeting the SLA. Monitor at successive levels of control (for example, the key IT layers: network, operating system, hardware, application, and so on) to make sure that the operating level requirements are met. AnalysisAs explained previously, the key IT layers can be helpful in identifying components that can be monitored and tuned:
In this phase, data monitored and collected is analyzed and used to carry out tuning exercises and establish profiles. These profiles are important since they allow the proper identification and adjustment of thresholds and alarms. When exception reports or alarms are raised, they need to be analyzed and reported upon, and corrective action needs to be taken. Ideally, all thresholds should be set below the level at which the resource is over-utilized or below the targets in the operating level agreement (OLA) or layered operating level requirements (OLRs). By doing this, capacity management can take corrective action before the targets in the OLAs have been breached or the resource has become over-utilized, causing a period of poor performance or failure. The organization’s service catalog should be reviewed for service capacity business impact analysis (BIA), ROI analysis, and capacity implications for IT service continuity planning, and as an initial baseline for workload-related issues and demand-management feedback. The data collected from the monitoring information should be analyzed to identify trends from which the normal utilization and service level, or baseline, can be established. By regularly monitoring and comparing this baseline with current resource usage, exception conditions in the utilization of individual components or service thresholds can be defined, and breaches or near misses in the OLAs can be reported. In addition, the data can be used to predict future resource use. Analysis of the data may identify the following issues:
The use of each resource and service must be considered over short-, medium-, and long-term periods, and the minimum, maximum, and average utilization during these periods must be recorded. Typically, the short-term period covers utilization over 24 hours, the medium-term period may cover from one to four weeks, and the long-term period covers a year or more. Over time, trends in the use of the resource by the various IT services become apparent. One key to determining whether a solution is operating at an acceptable level is latency, or the length of time a user has to wait for a response once a request for information is made. Heavy workload on a server might create unacceptable wait times even though the server may be capable of handling every request. As a rule, try to isolate components that have repeatable, high percentage contributions to performance levels and report them at varying workloads. It is important to understand the utilization levels in each of these periods so that changes in the use of any service can be related to predicted changes in the utilization level of individual resources. The ability to identify the specific hardware or software resources on which a particular IT service depends is improved greatly by an accurate, up-to-date, and comprehensive configuration management database (CMDB). Any relevant detail performance information should be maintained in the related but separate capacity management database (CDB). When the utilization of a particular resource is considered, it is important to understand both the total level of utilization and the utilization by individual services of the resource. ModelingModeling is a central element of the capacity management process. Modeling techniques and effective use of simulation software make it possible to investigate capacity planning “what-if” scenarios in order to build a model that simulates the desired outcome. Modeling relies on data from other capacity management subprocesses and activities, particularly forecasts of workloads and the use of resources by applications in development, and associated hardware specifications. Modeling should enable capacity (least cost) “value” decisions to be made from trend analysis. It can involve estimations based on past figures (trending), but it is much better and more accurate to make decisions on the results of techniques such as Transaction Cost Analysis, or simulation of the actual application. (For more information, see “Using Transaction Cost Analysis for Site Capacity Planning” at http://www.microsoft.com/technet/prodtechnol/comm/comm2002/plan/cs02tcas.mspx.) Figure 2 summarizes the modeling process and shows the three suggested stages to modeling: baselining the modeling environment, running “what-if” scenarios, and initiating changes to the services or resources being modeled. Analytical modeling, with algorithms based on IT service data, typically involves elaborate software that generates predictive and proactive capacity information. Simulation modeling is usually more accurate than analytical modeling since it provides greater capabilities and more closely emulates the application. However, simulation modeling may require significantly more time and cost to prepare scripts and generate realistic transaction loads. Specialized software that captures user interaction and plays back this activity at varying levels can make this type of modeling more cost-effective. Establishing a “baseline model” that accurately reflects the performance levels and workload characteristics of the current system is an important first step. This can allow predictive “what-if” scenarios with planned changes to resources under a variety of workloads. A good prediction of performance levels depends on the accuracy of the baseline model and its ability to accommodate and reflect changes required for the predictive model. Each “what-if” iteration generates several reports, which should be analyzed to see if the desired outcome has been achieved. Modeling, in its simplest and least expensive form, provides estimates based on experience and current resource utilization. This may be most practical for day-to-day small decisions. Pilot studies, prototypes, or full-scale benchmarks may be more appropriate for larger projects in which cost and time are justified. Certain architectural approaches to hardware and software deployment—for example, lower cost of pooled redundant storage—may be a method of gaining scalability without significant change to the environment. Effective pilot testing needs to include functional testing of as many variables as can be reasonably isolated for control. Testing needs to be persistent and constant over a specified time frame. There should also be a stress test, if feasible, to push the solution until it “breaks.” This helps in understanding the limiting resource components involved in the solution. Most capacity management activities should be run as projects and have an IT operations management sponsor with sign-off authority, who coordinates closely with the design and end-user community. OptimizingAnalysis of the monitored data may identify areas of the configuration that could be tuned to better use the system resource or improve the performance of the particular service. Optimizing techniques that can be helpful include:
Establish effective tuning guidelines that anticipate the growth by analyzing existing capacity information. Beware of supplier or direct vendor control of their package, hardware, or service. Especially, do not allow the vendor to tune any systems unless appropriately overseen by a qualified member of the IT organization or capacity management staff. To achieve acceptable transaction response time under maximum load conditions, tune the system to optimize system resources. Trend analysis of monitored data must also project anticipated growth trends. Systems should be designed and implemented with this growth in mind. During the design or building of a new application or system, there are often recommendations for tuning that should be examined. These recommendations are oftentimes available through application sizing tools. Flexible architectures are often better investments since they can expand or be tuned in the main areas of CPU, memory, disk or I/O speed, or network bandwidth. For example, a database transaction log file is written sequentially, often at the same time as data elements. By separating them, the log file on a disk is protected and optimized for write (for example, a separate mirrored pair), and the performance and availability of the database is improved. Likewise, the speed of striping many disks—for example, redundant arrays of independent disk (RAID) levels—can vastly improve the data and index performance and provide tuning options as well. It is a good practice to consult hardware and software platform white papers or technical notes for additional techniques to improve the design for capacity and tuning of the total solution after its deployment. As mentioned previously, these optimizing recommendations are typically obtained through the use of application sizing tools. Alternatively, a general-purpose modeling tool may be used to perform a what-if analysis. Change InitiationChange initiation introduces to the service in production any changes that have been identified by the analysis and tuning activities. This activity includes the identification of necessary change, and subsequent generation and approval of a change request. In some cases, implementation of the change runs concurrent to the service; in other cases, the type of change may require the service to be temporarily stopped. A slow, controlled pace for the entire change process is recommended. Reduce the number of variables in the implementation to improve the process. The implementation of any change should be done through formal change management processes. System tuning changes can have major implications for customer satisfaction. The impact and risk associated with these types of changes are likely to be greater than most other types of changes. Implementing tuning recommendations under strict change management procedures results in:
It is important that further monitoring takes place to assess the effects of the change. It may be necessary to make additional changes or to back out some of the original changes. Techniques of switching out an individual server (N+1) or a set of servers (N+N) to ensure smooth transition or “rolling updates” during change should be balanced with the cost. However, do not underestimate the cost of downtime, which can often justify this approach to implementing capacity change. In conclusion, effective implementation of capacity management requires careful analysis and a thorough understanding of the operation of the business, its processes, and the underlying causes for the use or non-use of capacity. Costs are assigned appropriately to the capacity categories according to the cause of capacity and type of cost. Monitoring, analysis, tuning, and change initiation (through change management) activities are iterative for both resource and performance management. For more information, see the MOF Job Scheduling, Storage Management, Service Monitoring and Control, and Network Administration SMF guides. Capacity Management SubprocessesIt is useful to consider capacity management as taking place at three levels within an organization: business capacity, service capacity, and resource capacity. Within the IT service management function, capacity planning is used as a tool to help decision makers optimize their choices, not just as a justification to buy more hardware. The Capacity Management SMF presents an end-to-end process that clearly demonstrates to users how to apply this, through the presentation of realistic business scenarios. Figure 3 shows all three capacity management subprocesses and how they interact with the CDB and capacity plan. Business Capacity ManagementAt the business capacity management (BCM) level, executives and other stakeholders evaluate current business, financial, economic, and technology indicators, with the goal of forecasting future business load placed on the IT system by business needs. This occurs at a corporate level, integrates with input from the other SMFs, and acts as input to the infrastructure engineering and enterprise architecture processes. The goal is to plan and implement projects to achieve necessary capacity on time and at an appropriate cost and risk. Scenarios at this level may include major changes to the business environment, such as business acquisitions or mergers. Inputs for BCM come from changes to a business plan, forecast, trend, or business need or through a periodic business review. The stakeholder review activity is where representatives from the business share and review each other’s high-level business plans. These plans are compared against the existing capacity plan and, if appropriate, the capacity plan is updated. A representative from IT should be present at this review to manage any major changes that need immediate action (using an RFC) and to update the CDB. The deliverable from this subprocess is an updated capacity plan. Service Capacity ManagementAt the service capacity management (SCM) level, the organization evaluates and bases short-term strategy and tactical response to service-related issues and needs. The goal is to profile and meet the needs of specific services on an end-to-end basis in order to remain in compliance with service level agreements. For example, recurring alerts for a certain set of events may indicate a technical issue that needs addressing, such as network congestion due to a faulty network switch, or may indicate a more process-oriented issue, such as unanticipated levels of online activity. Decisions at this level may be strategic or reactive in order to solve an immediate problem. Applicable scenarios may include the seasonal ramp-up of selected IT resources to maintain system availability (for example, to avoid excessive latency and user time-outs for an online toy store sales system in the two weeks before Christmas). The triggering events for service capacity management would come either from the change management process (as part of planning to implement a change to an existing service or to add a new service) or as the result of a known problem that affects service capacity. The service capacity management process may also recur periodically as a result of an operational management review. The outputs from this subprocess would be requests for change to the infrastructure or components as well as the updating of the capacity plan and CDB. Resource Capacity ManagementAt the resource capacity management (RCM) level, decision making typically occurs in a more reactive or tactical mode. The goal is to manage the capacity and performance of individual components within the infrastructure. Tasks initiated by capacity issues at this level may include hardware or monitoring optimization, service rescheduling, or other short-term or immediate actions. An example of an applicable scenario at this level would be bandwidth issues on a subnet where a department has just increased headcount and workstations. Inputs for SCM and RCM include the outputs from the Problem Management SMF that are attributed to capacity management issues, requests for change from change management that may affect capacity requirements, and the results of regular operations management reviews with capacity components. The outputs from this subprocess would be requests for change to the infrastructure or components as well as the updating of the capacity plan and CDB. Roles and ResponsibilitiesThe MOF Team Model organizes the activities of IT operations into seven distinct role clusters that represent areas, or functional roles, within IT operations (see diagram below). For more information on the Team Model, see the “MOF Team Model for Operations” white paper at http://www.microsoft.com/mof. Within the context of the MOF Team Model, the principal roles and their associated responsibilities for capacity management have been defined in this document according to industry best practice. This section outlines the roles and responsibilities of the capacity manager. The capacity manager role owns the capacity management process. It is important to note that this is a role, not a job description. In a small organization, one person may perform several roles, including capacity management, while a large organization may have a team of people designated just for capacity management. For example, a larger organization may have specialist coordinators and a staff of people assigned the job of carrying out the directives they receive from the capacity manager. It is recommended, however, that only one person perform the capacity manager role. Capacity ManagerThe capacity manager role is part of the Infrastructure Role Cluster within the MOF Team Model. (The Infrastructure Role Cluster is responsible for conducting long-term planning to ensure that plans are in place to meet the new and changing requirements of running the business from a networking, telecommunications, hardware, and software perspective.) The capacity manager role oversees the allocation and delivery of service capacity to users. The capacity manager is responsible for planning, monitoring, and reporting activities relating to system and solution capacity, performance measurement, and forecast in the IT organization. The capacity manager may have many associated activities, which often include:
The capacity manager is also responsible for managing the day-to-day capacity requirements of services, including:
Relationship to Other SMFsCapacity management is a service management function (SMF) in the Optimizing Quadrant of the Microsoft Operations Framework Process Model. It follows the MOF SLA Review milestone. The process is at a tactical level to assure quality of service for solutions offered by IT. Capacity management also manages the capacity change process, providing necessary design and planning feedback to planning activities described in the Microsoft Solutions Framework (MSF). Capacity management initiates ongoing improvements to IT service capacity and performance by delivering recommendations and management information plans and reports and by monitoring and controlling IT services. It also works to specify and refine requirements resulting from service level agreements (SLAs) that, in turn, drive operating level agreements (OLAs) internal to IT organizations. Capacity management’s core relationship is with service level management. It capitalizes on cost management for most of its recommendations and, for its normal optimization, is tightly coupled to availability management. Capacity management also provides input for contingency planning. For more information, see the MOF Service Level Management, Financial Management, Availability Management, and Service Continuity Management SMF guides. The capacity management process focuses on the appropriate capacity or performance levels of new technologies, existing applications, underlying system software and hardware, supporting tools, and associated documentation. Insufficient or degraded capacity can severely impact the availability of a service. Capacity and availability management functions often share common staff and management information tools. There is a need for coordination of related activities since they share similar management data views, vendor-supplied software, and a common goal of minimal disruption to a service due to change. The planning, decision-making criteria, and successful implementation of a recommended change required for capacity in the IT environment depends on the coordination of many related activities within the organization. MOF has service management functions (SMFs) that suggest strategies for improved coordination of IT operations. This section outlines the relationship of capacity management to these specific MOF SMFs. Infrastructure EngineeringInfrastructure engineering establishes a process by which the engineering and development groups within an organization communicate with the operations group, primarily to ensure that new applications or other acquisitions will be compatible and consistent with the existing infrastructure. This service management function also includes the development of infrastructure standards for a variety of IT and environmental components. Within the context of capacity management, adherence to the communication process and standards will ensure that new engineering projects will not adversely affect infrastructure capacity. Financial ManagementCapacity management creates upgrade plans that are included in the budgeting process. Accurate cost information is vital in order to accurately budget capacity upgrades. Planning for capacity management entails the planning for new hardware and software. These costs should be incorporated into the annual budget. Costs may be the restraining factor in some decisions and affect SLA negotiation. By effectively estimating the cost of service availability and optimizing capacity, IT weighs risk versus cost to identify the countermeasures they can afford to implement and those that should be reserved as contingency plan scenarios. Sometimes the return on investment for a requested change may need to be demonstrated. Capacity management must ensure that the necessary resources are acquired and implemented in a cost-effective manner. For more information, see the MOF Availability Management and Service Continuity Management SMF guides. Service Level ManagementAgreements between the provider and consumer of IT services exist within many organizations, yet in some they are unofficial, taken for granted, or unclear to one or both parties involved. Review and preparation of SLAs is a primary activity of service level management. Capacity management also helps define OLAs that result from service level requirements. IT must prioritize service alerts and countermeasures to prevent degradation of performance before it affects availability. Capacity management interacts closely with service level, availability, service continuity, and financial management staffs to decide on the cost-justified proactive measures required to improve the quality of service. Availability ManagementAvailability management ensures optimal availability of IT services through the correct use of resources, methods, and technology. Capacity management has a very close tie to this process, since optimal use of IT resources to meet performance levels at a justifiable cost relates to the goal of higher service availability. Shared reports should highlight trends indicating capacity or performance issues. Management information tools are typically leveraged to provide the monitoring information required for both processes. The availability plan needs to be coordinated with the capacity planning process. The same technology solution can often meet the needs of both plans. The availability and capacity plans should be created in collaboration. Some solutions that cannot be cost-justified for one plan may be justified in combination with the other. Service Continuity ManagementService continuity management copes with, and recovers from, unplanned situations in which the period of IT service disruption is considered unacceptable and normal availability countermeasures have not succeeded. There is a difference in scope of effect and the unacceptable nature of the disruption. Availability management deals more practically with what IT can effectively manage as part of its routine operation, but principles, approach, and concerns are similar. Both processes depend on input from capacity management to judge the level of performance when the countermeasures are enacted. Storage ManagementStorage management covers the day-to-day activities required to effectively operate and maintain storage management in an IT environment. It is a key SMF of the overall system administration process. Data backup, restore, and recovery operations are addressed. The SMF describes detailed components of capacity or performance monitoring. It presents specific aspects of Microsoft Windows® 2000 storage-related capacity and performance “how to.” Job SchedulingJob scheduling covers the day-to-day capacity-related activities involved in batch job processing. The job scheduling process is concerned with the details of batch job control, monitoring, and tuning. It is an example of capacity management practice applied to day-to-day operations. Capacity management sets requirements for job administration, workload, and performance expectations. Network AdministrationThis SMF provides network usage, bandwidth, and trend analysis to provide predictive capacity planning information accessible from a central facility that may populate the capacity management database (CDB). Capacity management draws on this data as input to provide demand management information and proactive recommendations. One of the functions of network administration is to provide tools that monitor network service usage. These tools help to quantify IT service capacity and availability. As components of capacity management, proper workload and demand management of network bandwidth rely on best practice network administration activities. System AdministrationSystem administration covers the ongoing operating activities required to implement change and effectively operate and maintain IT service and component system capacity and performance levels. The system administrator may make an approved capacity change; however, capacity recommendations need to be dictated by capacity management or job scheduling staff and the delegation controlled by their functional requirements. Change ManagementCapacity management assesses the impact of changes on existing capacity and identifies additional resource requirements based on the change in demand. Changes required for capacity management are implemented typically through planning and recommendations that result from capacity, availability, service continuity, and workforce management. Near-term daily operational changes may surface as job scheduling accommodates the more routine capacity changes. In all cases, any change to the IT service environment is channeled through change management as a request for change. Configuration ManagementChanges made to IT resources, also known as configuration items (CIs), and to service level objectives (SLOs) for these resources need to be reflected in the configuration management database (CMDB). Service level agreement availability and capacity data from the CMDB allows more proactive measurement of performance based on SLA compliance. This data is an important input to capacity management. Associated demand and workload requirements, resulting performance, and resource metrics are recorded in the capacity management database. Effective coordination and correlation of elements between these logical databases are required for timely information and ongoing capacity recommendation and planning. Problem ManagementProblem management deals with determining the root cause of problems. A problem is defined as one or more incidents exhibiting similar symptoms. Capacity management interacts with problem management to investigate known errors that have affected performance levels of an IT service. Capacity management also provides a specialist infrastructure role to identify and diagnose capacity- or performance-related problems. Capacity management provides ongoing feedback and recommended changes that result from incidents traced to known errors causing degraded performance levels of the service. Service DeskIncident frequency and statistics with respect to service performance levels may be reported through problem management, CMDB record, or involvement by the capacity management specialist to address known errors relating to performance and storage capacity. Ideally, IT resource performance is recorded and managed by the service desk and maintained by configuration management in the CMDB for historical retrieval and analysis. Workload, performance, and demand management activities may reference CMDB records resulting from incident escalation, failover and recovery capacity issues, or other tracked incident reports and trends. Service Monitoring and ControlThis process provides the foundation for determining service performance levels. Optimization of service performance implies monitoring the application’s end-to-end response times. In a well-run IT environment, performance levels are forecast and the monitoring system sets threshold alarms to trigger alerts before the customer of the service becomes aware of an issue. Service monitoring and control is central to alerts affecting capacity and availability, and for judging the need for contingency or service continuity measures. Operating level requirements for service monitoring require management data views through custom or selected vendor software products. AppendicesAppendix A: The Capacity PlanThe capacity plan should document the current and predicted levels of resource utilization and service performance. It needs to consider business strategy and include in its forecast of future requirements the resources that support the delivery of IT services or the planning of new ones. Any recommendations the plan makes should include quantified details of necessary resources, any relevant impact, and associated costs and benefits. Production of a capacity plan should occur at predefined intervals. Ideally, the plan should be published quarterly or in line with the business or budget life cycle. Updating the capacity plan should occur at predefined intervals or be triggered as the result of a significant change in business needs. A reissue of the updated plan will be necessary to account for changes in business plans, to report on the accuracy of forecasts, and to make or refine recommendations. This takes extra effort but, if it is regularly updated, the capacity plan is more likely to be accurate and to reflect the changes in business plans or requirements. A typical capacity plan will contain the following sections: IntroductionCompany background information should be introduced in this section. This can include the organization’ s current levels of capacity, current or anticipated problems due to overcapacity or undercapacity, the degree to which service levels are being achieved, and what has changed since the last update of the plan. Scope of the PlanIdeally, all IT services and resources need to be outlined in the plan. This section should specifically state which elements of the IT infrastructure are being addressed. Methods UsedThe capacity plan uses information gathered by subprocesses. This section should therefore contain details of how and when this information was obtained—for example, business forecasts obtained from business plans, workload forecasts obtained from users, or service level forecasts obtained by the use of modeling tools. Management SummaryThe capacity plan contains technical details that may not be of interest to all readers of the plan. The management summary needs to highlight the main issues, options, recommendations, and costs. It may be helpful to produce a separate executive summary document that contains the main points from each of the sections of the more detailed plan. Business ScenariosThe plan should consider and describe the current and future business environment. For example, a new customer relationship management (CRM) solution may currently be utilizing 60 percent of current processor and memory capacity for its back-end database. Capacity management is involved in monitoring the current system and is able to forecast the recommended additional CPU, memory, and disk capacity to accommodate growth for the year. It is important to explicitly mention all known business forecasts so that readers can determine what is inside and outside the scope of the plan. Service SummaryA service profile should be provided for each service delivered. This should include resource utilization for a given transaction response time or throughput rate. For example, usage levels for processor, memory, storage, and network, as well as short-, medium-, and long-term trends, are presented in this section. Forecasted Service LevelsThe business plans should provide the capacity manager with details of the new services planned and the planned growth or contraction of existing services. This section should report on new services and the removal of legacy or redundant systems. Resource SummaryThis section concentrates on the resulting resource usage by the services. It reports again on the short-, medium-, and long-term trends in resource usage, broken down by hardware platform. This information needs to be gathered and analyzed by other service management activities such as resource management and service performance and should therefore be readily available. Resource ForecastsThis section forecasts the likely resource usage resulting from the service forecasts. Each business scenario mentioned as part of the capacity plan should be addressed here. For example, a new Internet storefront project plan might include a forecast of network bandwidth requirements in anticipation of transaction levels and response time for the secured debit transaction. This forecast will use trending information gathered over a period of time to help determine and validate transaction levels based upon past experience and seasonal variation (peak usage times based upon the nature of the storefront). Options for Service ImprovementBuilding on the results of the previous section, this section outlines the possible options for improving the effectiveness and efficiency of service delivery. It might contain options for merging different services on a single processor, upgrading the network to take advantage of technological advances, tuning the use of resource or service performance, rewriting legacy systems, or purchasing new hardware or software. Cost ModelThe costs associated with the service option improvements should be documented here, as well as the current and forecasted cost of providing IT services. In practice, the capacity manager obtains much of this information from the financial management process. RecommendationsThe final section of the plan should contain a summary of the recommendations made in previous plans and their status. Any changes made to the previous plans or new recommendations should be documented here. The recommendations should quantify business benefits to be expected, potential impact of carrying out the recommendations, risks involved, resources required, and both startup and ongoing costs. Typical reports or capacity recommendations need to address the following areas:
Appendix B: The Capacity Management DatabaseThe capacity management database (CDB), not to be confused with the configuration management database (CMDB), contains the detailed technical, business, and service level management data needed to support the capacity management process. It is used for recording current resource and service performance level data necessary for trend analysis, as well as for forecasting and planning reports. Portions of the data may overlap, link, or synchronize with other process repositories—for example, the CMDB. The configuration management process is responsible for keeping configuration items up to date in the CMDB. The capacity management process, on the other hand, records performance information about those configuration items in the CDB. The following table represents an example layout for a typical CDB.
The CDB is a central repository and is used to store and exchange capacity and performance data between the three subprocesses of the Capacity Management SMF. In addition, the CDB can be used by a number of the other SMF processes, such as service monitoring, problem management, and change management. Inputs to consider for the CDB are:
Other detail input for the CDB includes:
Appendix C: Scenario-based ExampleThe following sections “act out” the three capacity management subprocesses for the fictitious scenario described in the following table. Each section describes the typical inputs and outputs and summarizes the process actions and objectives. The purpose of this example is not to teach capacity management; instead it aims to illustrate, using a simple example, how the three capacity management subprocesses work together to analyze different types of information and how they all contribute to the overall capacity management needs for this given scenario.
Business Capacity ManagementThe Woodgrove executive board and management teams have established a team to manage the acquisition of Contoso and coordinate all aspects of the transition phase. This acquisition team is made up of business unit managers, IT, operations, and finance. The acquisition team is responsible for managing the technical and business impact of the acquisition process to ensure a smooth transition and integration of the Contoso business and staff to Woodgrove. Inputs
Process SummaryThe acquisition plans are reviewed in detail with the stakeholders and a set of critical success factors for the acquisition is compiled. These success factors include logistical, financial, legal, operational, and business elements of the acquisition. Risk management is a key part of the decisions at this stage, so a detailed risk assessment is undertaken and a risk register is created. (For more information, see the “MOF Risk Management Discipline for Operations” white paper at http://www.microsoft.com/technet/solutionaccelerators/cits/mo/mof/mofrisk.mspx.)The business plan for the new combined business is also reviewed to ensure that the capacity plan is aligned with the strategic business needs as well as the short-term acquisition needs. OutputUpdated information in the capacity plan and CDB. Service Capacity ManagementThe MOF Service Role Cluster is responsible for ensuring that the messaging services being provided to customers are aligned to the customers’ stated needs. The Service Role Cluster maintains a working relationship with users and other stakeholders, understanding their need for messaging services and managing the service impact of the acquisition. InputsThe SLA for messaging describes the minimum levels of service for the existing (pre-acquisition) messaging system. The updated CDB and capacity plan needs to be carefully reviewed since there may be new demands on the messaging service as a result of the acquisition. For example, messaging may become more time critical between two departments in Chicago and Tokyo as a result of a new global advisory service offered by the group to corporate clients. Process SummaryThis process defines and agrees upon a new SLA for the combined (post-acquisition) messaging system, taking into account the new business needs as stated in the acquisition plan and business plan. This SLA should express the new levels of service in terms of critical success factors—that is, measures of what the business expects the messaging system to offer and key performance indicators that define the metrics and value by which the performance will be measured. Outputs
Resource Capacity ManagementHaving a clear understanding about the service capacity requirement for the new messaging system, the architects at Woodgrove need to design a new global messaging architecture for the new combined group. Inputs
Process SummaryUpdated information about the existing deployment and workload of Exchange is collected from the CMDB and monitoring system. A simple model is created showing the topology of the existing Exchange servers in the various branches of Woodgrove. Each Exchange server is represented as a black box, and lines representing the network connections connect the black boxes. This model is now validated to ensure that it is an accurate, complete, and up-to-date representation of the existing Exchange infrastructure. Once complete, this model is baselined and undergoes change management. The new Contoso locations are now added according to the proposed messaging architecture, including the network links. Since one of the stated needs was to re-use the existing hardware at Contoso, these have to be profiled and their capacity and performance parameters assigned to the black boxes that represent the servers at Contoso. For each possible Exchange architecture, a “what-if” simulation is run in order to generate Utilization, Transaction, and Bottleneck Analysis reports. Depending on the simulation method and technology used, reports can be generated to show how the model will perform for a given workload. The system used at Woodgrove generates the following reports:
A number of “what-if” simulations are now run. Each time, the reports generated are analyzed to ensure the model architecture performs within the SLA at the required capacity levels. For each simulation, it may be necessary to optimize the model architecture until it meets the performance, capacity, and cost needs. At the end of the modeling task, Woodgrove arrives at a messaging architecture that satisfies the business and service capacity needs. Outputs
Appendix D: Recommended TechnologiesAll organizations that intend to implement capacity management would benefit by obtaining and making use of certain tools and technologies. The appropriate number and complexity of these tools depend on the size of the organization and the number and type of IT components it wants to manage. This service management function guide takes a middle road by describing the tools needed to support the detailed processes that make up capacity management. The tools described here are sufficiently generic to enable all types and sizes of organizations to apply the advice. Several Microsoft tools can help with the capacity management process. These include:
Appendix E: Capacity Planning ChecklistThe following checklist describes a process that organizations may follow in completing or updating their capacity plan. This checklist was adapted from a more specific list developed and adopted by Microsoft’s MSN organization. Capacity Planning Process ChecklistThe MOF Capacity Management Service Management Function provides detailed guidance for implementing capacity management within an IT organization. This annotated checklist provides a simple means to track your organization’s progress in the development of a capacity plan, as described in the SMF.
Checklist adapted from “Capacity Planning Process and Plan Document Template” provided by Microsoft MSN Capacity Planning team.
|
In This Article
|
