Thought Leadership White Paper Holistic Capacity Management for Cloud and Virtual Environments Optimize service delivery by aligning compute resources with demand By John Seifert Product Marketing, BMC Software
Table of Contents 1 EXECUTIVE SUMMARY 2 THE VALUE OF CAPACITY PLANNING AND WHY PEOPLE LOSE SIGHT OF IT The cycle of IT capacity planning 3 WHY CAPACITY PLANNING MATTERS MORE THAN EVER The evolution of bimodal IT New cloud paradigms New models for business applications and shared IT services 4 THE COSTS OF NOT PLANNING FOR CAPACITY Too much capacity Too little capacity NEW REQUIREMENTS FOR CAPACITY PLANNING 5 IMPACT OF CAPACITY PLANNING ON SERVICE DELIVERY, BUSINESS AGILITY, AND COST 6 WHERE TO BEGIN CONCLUSION
Executive Summary Capacity management is the most underestimated problem of cloud computing, according to Morgan Stanley Executive Director for IT Strategy, Evangelos Kotsovinos. 1 IT organizations often think of capacity planning and management in terms of traditional IT environments, yet they are vitally important for today s virtualized systems and private and public clouds as well. As business applications become more complex and dynamic, IT must address capacity planning across the enterprise to ensure agile, high-quality service delivery at the lowest cost. This white paper discusses the importance of capacity planning and management for virtual, public, and private cloud environments, and how it can help businesses: Deliver digital services on time Plan and allocate the right compute resources for new applications and services when and as needed Scale up or down as needed Quickly adjust compute resources to meet peak, cyclic, and future service requirements Optimize costs Eliminate unplanned and unnecessary capital expenditures and invest in new compute resources only as required This white paper also gives guidance for a suitable starting point and next steps that will help IT organizations implement capacity optimization effectively within next-generation architectures, providing a flexible and cost-efficient foundation for digital service delivery. 1 Archana Venkatraman, Capacity management most underestimated cloud problem, ComputerWeekly.com, January 31, 2013. 1
THE VALUE OF CAPACITY PLANNING AND WHY PEOPLE LOSE SIGHT OF IT Capacity planning and management have a vital role to play in helping IT provide optimal support to the business as efficiently as possible. Beyond maximizing the use of existing compute resources, an effective planning capability can help IT understand the impact on capacity of business and technology roadmaps, prevent capacity issues from delaying business projects and services, and seamlessly accommodate organic growth in service demand. Accurate information about future capacity needs also equips IT to negotiate deals to buy the right amount of infrastructure at the right time, at the right price, instead of relying on smaller, reactive purchases on less favorable terms. People tend to think about capacity planning and management only when forced to by economic pressures, and lose focus at other times, such as when prevailing enterprise technologies become cheaper. But less-costly technologies also generally lead to increased consumption and more-complex models, making management more difficult and sowing the seeds for future problems as the need for capacity planning reappears later in the cycle. This trend has played out repeatedly over recent decades. The cycle of IT capacity planning During the 1970s and 1980s, mainframes and other extremely expensive technologies played foundational roles in enterprise IT, making optimization essential. As a result, centralized IT and capacity planning became widespread. By the late 1980s and into the 1990s, though, a rapid shift from $4 million mainframes to $2,000 servers made this capacity planning seem less urgent, while a concurrent move toward decentralization brought a less-holistic approach to IT in general. As IT groups and business units took advantage of easily procurable, low-cost hardware to meet their needs, enterprises ended up with excess capacity throughout their environment, but didn t pay much of a penalty for this inefficiency. Perhaps understandably, capacity planning fell out of favor. The surge in internet technologies leading up to the year 2000 and beyond saw the rise of larger, more-powerful servers that were also much more expensive, running $500,000 or more. At the same time, the need to address the much anticipated Y2K bug enterprise-wide brought a wave of IT re-centralization. Once again, capacity planning moved to the forefront of IT management. In recent years, as virtualization and cloud technologies transform enterprise environments and enable low cost capacity on demand, some IT organizations are taking their eyes off capacity planningbut this is a mistake. According to IDC, The ability to track cloud resource consumption to support life-cycle management, capacity planning, and (optionally) chargeback/showback is one of six basic tenets of cloud system management. 2 RAPID GROWTH ADDS URGENCY With cloud adoption expected to expand three-fold or more over the coming four years, a better approach to capacity planning and management is becoming a top priority for IT organizations. Worldwide Cloud Systems Management Software Revenue by Segment, 2014 and 2019 Private/hybrid cloud Public cloud 0 1,000 2,000 3,000 4,000 5,000 2014 2019 IDC predicts rapid cloud adoption over next five years. 3 2 IDC Worldwide Cloud Systems Management Software 2015 2019 Forecast, April 24, 2015. 3 ibid 2
WHY CAPACITY PLANNING MATTERS MORE THAN EVER Three trends make capacity planning especially relevant and important for today s businesses: 1. The evolution of bimodal IT 2. New cloud paradigms 3. New models for business applications and IT services that support them The evolution of bimodal IT As defined by Gartner: Bimodal IT refers to having two modes of IT, each designed to develop and deliver information- and technology-intensive services in its own way. Mode 1 is traditional, emphasizing scalability, efficiency, safety and accuracy. Mode 2 is nonsequential, emphasizing agility and speed. 4 IT is accustomed to thinking of capacity planning in terms of Mode 1, with a focus on traditional environments. But capacity planning is also vital for Mode 2 because it helps IT make effective strategic use of virtualized and cloud resources to deliver high-quality digital services with optimal agility, flexibility, and efficiency. This dualistic mode of operations, combined with the growing complexity and diversity of enterprise environments, places renewed importance on strong operational practicesand makes holistic capacity planning essential for effective management. New cloud paradigms Seeking to increase scalability and agility while controlling costs, more CIOs are incorporating private and public clouds into their IT infrastructure. Capacity planning and management have a vital role to play in cloud strategy; as Evangelos Kotsovinos said at the Cloud Expo Europe 2013 event, One of the main reasons for using cloud computing services is to get efficiency and cost savings. And maximum IT efficiency on the cloud comes from good capacity planning and management. 5 But the trends shaping the adoption of cloud in the enterprise also make it more challenging to achieve capacity optimization. 4 Gartner, Gartner IT Glossary. www.gartner.com/it-glossary-bimodal. Accessed August 2015. 5 Archana Venkatraman, Capacity management most underestimated cloud problem, ComputerWeekly.com, January 31, 2013. 3
Workloads Moving to Private Cloud What percentage of your workloads do you expect to be on a private cloud in the next 12 months? More than 75% 11% Between 50% and 75% 11% Between 25% and 50% 19% Between 10% and 25% 32% Less than 10% 26% Source: BMC December 2014 Capacity Management Survey. Most enterprises expect to shift a large percentage of workloads to a private cloud during 2015. Although centralization continues to be an important element of IT strategy, with many applications and other services delivered on demand from the data center, the ready availability of cloud services also encourages individual IT, application, and business teams to source and leverage their own cloud resources. These teams increasingly develop and manage their own service models and perform capacity planning within their own organizations, adding to the complexity of the overall enterprise environment. As a result, it is more difficult for IT to control cost and ensure agile, high-quality service delivery across the enterprise. New models for business applications and shared IT services The applications that provide business services today are more complex than ever, each perhaps encompassing nine tiers and hundreds of servers. These applications undergo both vertical growthrunning on more-powerful servers to speed processingand horizontal growth, where applications are spread across additional business tiers to support a larger number of concurrent users. Meanwhile, the IT service tiers that support business applications, including database, Java, OS, web, and so on, are aligned to infrastructure teams that perform some level of capacity planning for their own services, but not the enterprise-wide management needed for effective optimization. All of these new and evolving models lead to greater management complexity, with fewer skilled personnel addressing capacity planning in a holistic and centralized way, and more people in different groups addressing only their own local needswithout being connected to each other through effective enterprise workflows. THE COSTS OF NOT PLANNING FOR CAPACITY As illustrated above, capacity management is both more important and more difficult than everand yet few organizations are doing it and even fewer are doing it well. Too much capacity Too many IT groups use spreadsheets to estimate capacity, a manual, clunky, time-consuming, and error-prone approach. Without a precise way to anticipate requirements, IT buys more hardware than needed just to be safe, straining capital and expense budgets. Similarly, overprovisioning virtual resources contributes to sprawl and management overhead, along with acquisition of the wrong class of cloud services or acquiring services for overextended periods. Companies wishing to track the cost of services may find that the spreadsheet model makes accurate chargeback unattainable. Too little capacity At the other end of the scale, the inability to plan ahead leads to capacity issues that delay business projects and services. Undersourced capacity can lead to service delays or failures that affect customer-facing services. All of this undermines the agility so many businesses rely on for competitive advantage. 4
NEW REQUIREMENTS FOR CAPACITY PLANNING IT needs a new approach to capacity planning designed for the way businesses operate now. Cost and risk are still considerations, but more important is IT s ability to support agile development and push code out quickly to capture market share. With this development model, business units need to be able to deploy IT resources, functionality, and services at a speed that past capacity planning hasn t allowed. No longer just a control mechanism, capacity planning is now a tool for helping IT make dynamic changes to its infrastructure that help the business succeed, ensure scalability, meet shifting requirements, and realize the benefits of agile development. While the fundamentals of capacity planning and management remain the same, the pace and complexity of this function are now unprecedented. A complex and dynamic array of data center and cloud resources must be synchronized in real time to support service models throughout the business. In the past, a business unit facing a peak demand period might have appropriated available capacity from another groupa reactive, ad hoc mode of operations that can lead to chaotic waves of migrations to and from cloud resources. IT needs a more rational, holistic approach to view and address current enterprise demand optimally and in real time, making smaller incremental adjustments to synchronized cloud and virtualized resources on both sides of the firewall. IT also needs to anticipate requirements and understand how best to meet future demand at the best possible cost while fulfilling expectations for performance, scalability, and compliance. IMPACT OF CAPACITY PLANNING ON SERVICE DELIVERY, BUSINESS AGILITY, AND COST Effective capacity planning and management can deliver powerful business value in the cloud era. With the ability to visualize capacity needs across the enterprise, IT can accurately allocate compute resources to: Deliver digital services on time Plan and allocate the right compute resources for new applications and services when and as needed, and respond quickly to new business requests Scale up or down as needed Quickly adjust compute resources to meet peak, cyclic, and future service demands driven by organic growth, as well as accommodate the capacity impact of business and technology roadmaps Optimize costs Eliminate unplanned and unnecessary capital expenditures by maximizing use of existing compute resources, and invest in new resources only as needed and based on accurate information for negotiating large procurement deals Drivers for Capacity Management Which are the most important high-level drivers for the Capacity Management process at your company? Please assign a priority for up to three (1 = most important). Cost Reduction 56% 27% 17% Risk Avoidance 45% 38% 16% Alignment of IT to Business 26% 38% 36% Audit/Compliance 21% 30% 49% Cost Visibility 13% 33% 54% Business Agility 13% 38% 49% Chargeback 8% 23% 69% 1 2 3 Source: BMC Dec. 2014 Capacity Survey Enterprises cite a variety of high-level drivers for capacity management processes. 5
WHERE TO BEGIN As companies expand their use of cloud computing, business owners and service owners alike need visibility of services across the enterprise to ensure the performance and availability of digital services. Developing service models is also an important building block of capacity planning, providing insight into the relationships between specific resources and the services they power. Another key step is architecting applications that can grow horizontally, not just vertically, so capacity can be added easily just by provisioning a virtual machine (VM) in a public or private cloud. This makes it possible to start performing capacity planning for: Optimal utilization across private and public cloud resources Placement of workloads to run correctly at the right level of performance while complying with regulatory requirements Cost management by ensuring that workloads are being run on the most cost-effective resources and by using available capacity before paying for more CONCLUSION As businesses quickly leverage cloud resources for increased agility and scalability, organizations often overlook the importance of capacity planning and management to deliver the full benefits of this strategy. While cloud technology can greatly improve efficiency and reduce cost, these benefits aren t automatic. Without holistic management and optimization, the enterprise environment will grow increasingly complex and difficult to manage, undermining IT s ability to ensure high performance and availability of business services. Today s cloud-centric IT environments call for a new approach where capacity management acts not simply as a control mechanism, but also accelerates and strengthens the business. By gaining a holistic perspective on service models, private and public cloud resources, and demand patterns across the enterprise, IT can help the business: Deliver digital services on time to support business agility Scale up or down as needed to ensure high quality now and in the future Optimize costs by making the right investments in the right way, at the right time FOR MORE INFORMATION To learn more about capacity planning and management, please visit bmc.com/it-solutions/capacityoptimization.html ABOUT THE AUTHOR John Seifert has over 15 years of experience in capacity and performance management. John has been a capacity planner for mainframes, UNIX, and virtualized environments at CIGNA, UHG, and General Dynamics. He has also been a product manager for performance and monitoring solutions at TeamQuest and Candle. John is currently a product manager for TrueSight Capacity Optimization at BMC Software. BMC BMC is a global delivers leader software in software solutions solutions that help that IT transform help IT transform digital enterprises traditional for businesses the ultimate into competitive digital enterprises business for the ultimate competitive advantage. advantage. We have Our worked Digital with Enterprise thousands Management of leading companies set of IT solutions to create is and designed deliver to powerful make digital IT management business fast, seamless, and optimized. services. From From mainframe mainframe to mobile to cloud to cloud to mobile, and beyond, we pair high-speed we pair high-speed digital innovation digital innovation with robust with robust IT industrialization IT industrializationallowing our customers allowing our to provide customers intuitive to provide user experiences amazing user with experiences optimized with performance, optimized cost, IT performance, compliance, and cost, productivity. compliance, BMC and solutions serve more productivity. than 15,000 We customers believe worldwide that technology including is the 82 percent heart of of every the Fortune business, 500 and. that IT drives business to the digital age. BMC Bring IT to Life. BMC Bring IT to Life BMC, BMC Software, the BMC logo, and the BMC Software logo, and all other BMC Software product and service names are owned by BMC Software, Inc. and are registered or pending registration in the US Patent and Trademark Office or in the trademark offices of other countries. All other trademarks belong to their respective companies. Copyright 2015 BMC Software, Inc. *468079*