Data Center & IT Infrastructure Optimization Trends & Best Practices Mickey Iqbal - IBM Distinguished Engineer IBM Global Technology Services
IT Organizations are Challenged by a Set of Operational Issues Challenges Costs & Service Delivery Business Resiliency & Security Rising costs of systems and networking operations Explosion in volume of data and information Enterprises report that IT Difficulty in deploying new applications and services operational overhead = up to 70% Security of your assets & your clients information of IT budget and growing... leaving precious few resources for Landslide of compliance requirements new initiatives. Systems and applications need to be available Energy Requirements Rising energy costs & rising energy demand Power & thermal issues inhibit operations Environmental compliance & governance mandates Slide 2 February 2009
Project Big Green IBM to reallocate $1 billion each year: To accelerate green technologies and services To offer a roadmap for clients to address the IT energy crisis while leveraging IBM hardware, software, services, research, and financing teams To create a global green team of almost 1,000 energy efficiency specialists from across IBM Re-affirming a long standing commitment at IBM: Energy conservation efforts from 1990 2005 have resulted in a 40% reduction in CO2 emissions and a quarter billion dollars of energy savings Annually invest $100M in infrastructure to support remanufacturing and recycling best practices Invest in data centers we own or manage for clients to double compute capacity by 2010 without increasing power consumption or carbon footprint saving 5 billion kilowatt hours per year... equals energy consumed by Paris the City of Lights What green solutions can mean for clients: For the typical 25,000 square foot data center that spends $2.6 million in power annually, energy costs can be cut in half Equals the reduction of emissions from taking 1,300 automobiles off of the road Slide 3 February 2009
Addressing Energy Challenges in Data Centers Face similar issues to clients in supporting 8 million square feet of data center space Increasing computing demand Need to respond to new business growth Changing Cost Dynamics Increase in Electricity Prices IBM Data Centers (2005-06) Long lead times ranging from 12-18 months for major data center electrical and mechanical components to support IT growth Changing cost dynamics Latest generations of data processing equipment drive increased power and cooling demands Increased use of blade servers driving higher power and cooling requirements Percentage Increase 20 18 16 14 12 10 8 6 4 2 0 14 Global 18 US Clients experiencing similar utility infrastructure constraints, facing multi-million dollar upgrades IBM Data Centers Comparison of Space to Energy Use Aging data center infrastructure Building equipment reaching end of life Several sites converted from manufacturing to data centre Some locations unable to handle growth requirements due to physical constraints Percentage of Total 35 30 25 20 15 10 5 0 6 30 Space Energy Slide 4 February 2009
IBM s Data Centers IT Infrastructure Energy Efficiency Strategy Migrate many applications into fewer images Simplify IT environment Reduction of operations resources Improve application specific monitoring and tuning Consolidate many centers into fewer Reduce infrastructure complexity Improve facilities management Reduce staffing requirements Improve business resilience (manage fewer things better) Improve operational costs Centralization Consolidate many servers into fewer on physical resource boundaries Reduce system management complexity Reduce physical footprints Physical Consolidation Improved Operations Remove physical resource boundaries Increased hardware utilization Allocate less than physical boundary Reduce software licensing costs Virtualization Best Practices Application Integration State-of- the-art Hot and cold aisles Improved efficiency transformers, UPS, chillers, fans, and pumps Free cooling Integrated power management Direct liquid cooling Combined heat and power Conservation techniques Infrastructure energy efficiency Improved airflow management Facility Infrastructure Energy Efficiency Strategy Slide 5 February 2009
IBM s Data Center Energy Efficiency History A decade of improvement IBM Strategic Delivery Model IBM Metrics 1997 Today CIOs 128 1 TECHNOLOGY Host data centers 155 7 Web hosting centers 80 5 Network 31 1 Global Resources Strategic IGA Location Strategic Web Location for IGA Ethernet & Power9 Networks Applications 15,000 4,700 Tactical and operational efficiencies Consolidation of infrastructure Application consolidation/reduction Global resource deployment Enterprise end-to-end architecture optimization Slide 6 February 2009
Five Areas for an Energy Efficient Data Center Build Diagnose Get the facts to understand your energy use and opportunities for improvement Manage & Measure Data Centre Energy Efficiency Assessment Mobile Measurement Technology (MMT) Data Centre Upgrade Scalable Modular Data Centre Plan, build, and upgrade to energy efficient data centers IBM Project Big Green Seize control with power management software IBM Energy Management Solution Virtualize Cool Distributed Systems Virtualization - Advanced Power Virtualization (APV) and VMWare Linux on z Implement virtualization and other innovative technologies Use innovative cooling solutions Data Centre Stored Cooling Rear Door Heat Exchanger Slide 7 February 2009
Where does the energy go? The data center energy challenge affects both the physical data center & the IT infrastructure. % of total data center electricity use Server and storage consolidation assessments Data Center Energy Efficiency Assessment 35 Cooling systems Electrical and building systems 30 25 20 15 10 5 0 Information technology Chiller/ cooling tower Humidifier Computer room airconditioner Power distribution unit Uninterruptible power supply (UPS) Switch/ gen Lighting Power use Optimize IT Infrastructure Energy Efficiency Active Energy Management Chart and data source: American Power Conversion Corporation (APC) white paper, Implementing Energy Efficient Data Centers, #114, by Neil Rasmussen, 2006. Optimize Data Center Infrastructure Energy Efficiency Slide 8 February 2009
Build - IBM s Data Center Family TM Comprehensive Set of Custom and Standard Capabilities Scalable modular data center Turnkey data center for 500-2,500 sq ft Rapid deployment in 8-12 weeks 20% less cost than traditional data centers 15-30% improved energy efficiency Enterprise modular data center Standardized design in 5K up to 20K sq feet Designed for high availability Leadership energy efficiency 66% DCiE 25% faster deployment than custom approach Open architecture involving leading vendors Portable modular data center Fully functional data center; multi-vendor support Portable - temporary and remote data centers Rapidly deploy in 12-14 weeks Designed for high availability Leadership energy efficiency: 77% DCiE High density zone Plug and play infrastructure to support high density servers in existing data centers Non-disruptive implementation 35% lower cost than retrofitting existing data center Slide 9 February 2009
Measure and Manage Seize control of every resource with energy management software Trending consumption on individual or group level Establish baseline cost Retrieve temperature and power information Better utilization of existing resources Data Center Infrastructure Assets Tivoli Green Management (Monitor, Measure and Manage) Tivoli Software IBM Systems Director and Active Energy Manager IT Assets Facility Infrastructure Assets 3 rd Party Servers and Storage Slide 10 February 2009
Consolidation: four basic strategies Centralization New York Reducing the number of data centers Physical Consolidation Reducing older servers with fewer more powerful systems Chicago Chicago London Data Integration Centralizes data from different sources Application Integration Migrates several applications into fewer applications Windows NT Servers UNIX Servers Other Servers Windows NT Servers UNIX Servers Other Servers Billing System Server Web Server Order Server Billing Server Order Server Web Server Slide 11 February 2009 10/22/2009
Virtualization is central to data center transformation Fragmented, inefficient islands of computing Efficient, dynamic and responsive What Can You Do? Support your company growth Improve system performance & availability Reduce your power & cooling consumption Manage more servers Enable application deployment Free up money from operations Slide 12 February 2009 10/22/2009 The payoff with no increase in operational resources with far less hardware by 60% or more with a smaller staff in days versus weeks or months to fund IT projects that increase business revenue
Virtualization changes everything Requires enhanced management framework What happens when you virtualize? With the right tools, virtualized resources can be easier to create, adjust, move, clone and checkpoint However, new complexities can emerge Rapid growth of virtualized resources Relationship of virtualized resources to underlying physical infrastructure Health monitoring and problem determination across a physical and virtualized infrastructure Manage a mix of virtual and physical resources Reduce complexity with integration and service management Deploy and manage virtual servers to control datacenter space Manage energy helps meet green initiatives and cost objectives Slide 13 February 2009 10/22/2009
Computing cloud - an emerging IT infrastructure Delivers simplified services through innovative business models Clients and Customers INNOVATIVE BUSINESS MODELS SIMPLIFIED SERVICES Network Cloud Computing Cloud New combinations of services to form differentiating value propositions at lower costs in shorter time Cloud applications enable the simplification of complex services A cloud computing platform combines modular components on a service oriented architecture Internet protocol based convergence of networks and devices A high performance pool of virtualized computer resources Slide 14 February 2009 10/22/2009 14
Lessons Learned Plan for the long term knowing that your environment is not homogeneous or static Significant data center changes take years (not months) to implement Plan modularity for unknown future requirements Leverage the base capability you have already paid for Tactical increases in power density Don t underestimate the value of existing infrastructure Leverage new technology benefits in facilities and IT for function and efficiency Examples of equipment installed Immediate opportunities exist to reduce data center energy use Mobile Measurement Technology can show how to save 10% by simple actions Data Center Energy Efficiency Assessment allows you to do a before and after view Server/Storage consolidation offers large opportunity Slide 15 February 2009