May 2009
Enterprise Data Centre
The Challenges and Changes for the Data Centre
Manager
Introduction
Increase awareness of IT energy usage
Understand processes and practices available for identifying energy
usage
Data Centre issues
IBM and Green Energy initiatives
The Energy Problem
Energy provision has emerged as arguably the most
important business, geo-political and societal issue of our
time.
The organization that responds (with not only technology
innovation but also perhaps legislation to support the uptake
and application of technology) will create and take
advantage of this business opportunity.
IBM is well positioned to provide the services and
Power and cooling exceeds server
Spending – IDC
“Power and cooling will be a top 3
issue with all CIO’s in the next
6-12 months”
Michael Bell – Gartner Group
“Power and cooling costs will
increase to more than one-third
of the total IT budget”
Robert Frances Group,
January 2007
What is the greatest facility problem
with your primary data centre?
Gartner 2006 0% 5% 10% 15% 20% 25% 30% 35% None of the above
Poor location Excessive facility cost Insufficient raised floor Insufficient power Excessive heat $0 $10 $20 $30 $40 $50 $60 $70 $80 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 (M units) 2 4 6 8 10 12 14 16 18 Spending
New server spending
Installed base
(US$B)
Where does the energy go? The data centre energy challenge affects
both the physical data center and the IT infrastructure
Power Use 35 30 25 20 15 10 5 0 Chiller/ Cooling Tower Information
Technology UninterruptiblePower Switch/ Gen Lighting Supply Power Distribution Unit Computer Room Air Conditioner Humidifier
Cooling systems Electrical and building systems
%
of total data cent
re ele
c
tricit
y
Comparison of typical server utilisation rates
UNIX
x86
Mainframe
Facing new pressures: Data Centres are at a tipping point
“According to Gartner, by 2008, 50 percent of current data centers will have insufficient power and cooling capacity to meet the demands of high density equipment." 1
1. Gartner, Data Centre Power and Cooling Challenge, Gartner 25th annual data center conference presentation, Michael A. Bell, December 2006.
Facing new pressures: Data centers are at a tipping point
Increased Computing
Demand
Changing Cost
Dynamics
Data Center
Lifecycle Mismatch
US commercial electrical costs increased by 10
percent from 2005-06. - EPA Monthly Forecast, 2007
Datacenters have doubled their energy use in the
past five years.- Koomey, February 2007
Per square foot, annual data center energy costs
are 10 to 30 times more than those of a typical office building. - William Tschudi, March 2006
New low cost technologies like blades consume
20-30kW of power per rack, while the average data center is designed to support 2-3kW per rack
“Eighty-six percent of data centers were built
before 2001”2
In the past decade, growth in server shipments has
been 6x and 69x for storage - Bill Zeitler, IBM STG GM
1. Gartner, Data Center Power and Cooling Scenario Through 2015, Rakesh Kumar, March 2007.
According to Gartner, “The underlying consumption of energy in large data centers to power and cool hardware infrastructure is likely to increase steadily during the next ten years.”1
The Vision
an evolutionary new model for efficient IT service
delivery
Rigid and Complex
Sprawl / distributed
Inefficient power and
cooling
Siloed
9
Efficient, shared and highly virtualized
infrastructure
9
Automated service provisioning
9
Elasticity of scale
9
Responsive and business aligned
Fragmented, inefficient
islands of computing
Project Big Green
Project Big Green is IBM’s integrated response to carbon concerns. A comprehensive
initiative across all pillars of operations, offerings and policy. Reallocating $1 billion each
year – Reaffirming a long standing commitment at IBM.
Collaboration
Collaboration
Across
Across
Eco
Eco--systemsystem
Energy Energy Efficient Efficient Infrastructure Infrastructure
IBM Project Big Green
IBM Project Big Green
Five building blocks provide the approach to
operational savings and business growth.
Diagnose
Get the facts to understand your energy use and
opportunities for improvement
Build
Plan, build, and upgrade to energy efficient data centers
Cool
Virtualize
Implement virtualization and other innovative technologies Manage & Measure Seize control with energy management softwareIBM Project Big
Green
Use innovative cooling
Ability to accurately view baseline energy cost
Cost savings from more efficient energy use Rising global energy prices
Squeeze on IT budgets
Meaningful energy conservation and reduced
Positive impact linking corporate social More computing performance per kilowatt
More energy used for processing
Corporate social responsibility
Limited “green” public image Under utilised server systems
Exploding power & cooling costs
Financial
responsibility and personal values
Operational
Environmental
Benefits of a green data Centre
CO2 emissions
Data Centre Thermal Measurement and Analysis
Utilising IBM Mobile Measurement Technology (MMT)
Jointly developed between IBM and Pacific Gas + Electric
Announced one year ago as research product as part of Big Green
Launching as official service offering within IBM Global Services Site and Facilities
Thirteen assessments performed this past year
12,991,080 (kWh) energy savings opportunity per year
Average of 12% savings in IT power
Average of 23% savings in IT cooling costs
IBM Heat eXchanger - Removes heat at the source
After
Before
(between applications)
Virtualised Data Centre
• With Virtualised server and storage pools, peaks in demand by one application can be
accommodated with the pooled resources
• A Utilization improvement from 15% to 45% reduces power footprint by over 60%
HP Internet WAN LAN
CRM
ERP
SCM
VPN
Virtual Server
Virtual Storage
CRM
ERP
SCM
With silos for every application, designers
are required to meet the peak demand within
the server infrastructure for that application
If Virtualisation is so great… why isn’t everybody doing it?
Inadequate change management policies for such
a dynamic environment
Ensuring compliance and performance is not easy
with manual systems
Virtualization
Servers Network Storage
Physical Infrastructure Virtual Infrastructure