DC Service Management

Transcription

DC Service Management
Smarter Data Centers
Integrated Data Center Service Management
Approx. 3.500 ton CO² is generated by:
Diesel Car-6l /100 km –20.000 km
Airplane –5 * Brussels-Barcelona
Server, 1.5 A, 24*7*365*3.000 Kwhr
Benchmarks – Power Usage Effectiveness
PUE – Power Usage Effectiveness = Total Facility Power/IT Equipment Power
The Total Facility Power is defined as the power measured at the utility meter and is power
dedicated solely to the data center
IT Equipment Power includes servers, storage, and network equipment, plus equipment such as
KVM switches, monitors, and workstations/laptops used to monitor or otherwise control the data
center
Most energy
efficient
1.0
1.5
Objective
2.0
Current
Least energy
efficient
2.5
3.0
(1) Standard supported by Green Grid, originally published by Lawrence Berkley National Laboratory, 2003
Extend the life with improved data center energy efficiency
Address both the IT and physical infrastructure uses of energyto Start
% of total data center
electricity use
Electrical and building systems
Electrical and Building Systems
Cooling 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
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
Design new infrastructure to be responsive to change
Data center capital costs
Data center operating costs
60% costs from energy related components
75% costs from energy use
Mechanical
20%
Instrumentation
Fees
& Controls
24%
4%
Power
36%
Fit out
costs
9%
Shell
7%
Source: IBM Estimates
Energy efficiency provides an opportunity
to address operational efficiency, flexibility, and resilience
Extend the life of an
existing data center
infrastructure
Rationalize the data
center infrastructure
across the company
Double IT capacity
Or
Reduce operational
Expenses by 50%
Improve operational
Efficiencies while
Reducing operational
Expenses by 50%
Design new
infrastructure to be
responsive to
change
Pay as you grow by
Deferring 40 – 50% of
Capital and operational
costs
Where are all those kW consumed ?
100
80
Server
Hardware
Data Center
Server
Loads
60
40
20
0
Typical Utilization
Mainframe 80 – 90%
Unix
10 – 20%
Wintel
5 – 12%
PUE 1.65
AC/DC
losses; 25%
Fans; 9%
Used Resource
Idle;
20%
80%
HVAC,
UPS
40%
IT
power
60%
= 3,6% of total
Processor
DC/DC
losses; 10%
30%
Standby; 2%
Drives; 6%
PCI; 3%
Planar; 4%
Memory; 11%
Need one more W of
compute energy ?
27 W data center
X1,7
16 W equiv IT power
x3
Sites & Facilities
5.6W
x1.7
5 W equiv processor
x5
+1 W equiv. used ressource
Systems Design
3.3W
x3
1.1W
x1.1 @90%!
Who is responsible for energy efficiency in data centers?
IBM recommends that one person should be named as
being responsible for energy efficiency within data
centers
•
•
•
•
•
•
One person
One (virtual) organisation
One budget
One process
One set of guidelines and best practices
One set of tools
To manage energy efficiency you need to measure
and monitor energy consumption across the Data
Center ref European Data Center Code of Conduct
How is a Data Center Managed today
IT
Facilities
Data Center
How should a Data Center be managed
End to end Energy, Quality, Service and Risk Management
Data Center
Integrated Governance,
Service & Process Mgt
+ Tools
Data Center Integrated Monitoring and Management
Services (DCIMMS)
Simulation
Virtualize
Collect & Measure
Data Center
Management
Control
Manage
Integrated Management approach
Define and agree a common strategy and focus
related to Data Center services?
Determine gap matrix between IT and Facilities
processes
Define clear ownership for all major elements ie
energy
Start managing and monitoring energy across the
Data Center
Develop Data Center SLA’s and KPI’s?
Prioritize process improvements to reduce Data
Center risks and cost and improve quality of service
Data Center Service Management:
Delivering Quality and Cost effective Services
Visibility: See
your Business
Control: Manage
your Business
DC Service Management
DC Best Practices, Methodologies & Services
DC Service Management Platform
Automation:
Improve your
Business
Establish an open-based platform
Data Center Service Management
Data Center Best Practices, Methodologies, and Services
Data Center Service Management Platform
Do you need
to manage all
your key
processes
from a single
interface?
Does the
health &
performance
of your
applications
support your
business?
Visibility
Do you need
to manage all
of your data,
information
and storage
resources?
Are your
people,
applications
and data
secure?
Are you sure?
Control
Do you need an
energy-efficient
datacenter?
How will you
manage SOA?
Can you
visualize &
control of all
your assets?
Do you want
your asset
lifecycle
automated?
Can you
deliver
effective
network
services?
Is QoS
important?
Automation
Perspective on the data center monitoring and management
market
The current market is a fragmented one, with no standards. There are a myriad of suppliers
and solutions, but no single solution that bridges all user pain points.
Lacking
visibility
Clients lack true
understanding of data
center operational
issues:
IT/Facilities
Room for
growth
Clients
seeking
Efficiency
Understanding
need to manage
visibility to data:
Integrated IT and
Facilities ops
Time (years)
Gartner Group predicts 60x growth in this space through 2014
Manual
processes
Estimated
usages
Niche of
tools
Integration
IT and
Facilities
and to fully
optimize business
applications
operations
Right-sized facilities for optimized IT
Start improving data center efficiency
by monitoring asset utilization
Optimize utilization to gain power
savings and the opportunity to
reduce cooling
See efficiency results with
industry standard energy metrics
Ensure high availability by monitoring
temperatures in high density IT areas
for compliance
Use detailed reports to provide
energy savings and carbon
reduction for incentives €€
Overall CPU cycle
usage very low
Realize 60% of servers can be
running at less than 5% utilization
Drill down capability
identifies assets
with low utilization
Thank you
Graham Aldridge
Benelux Data Centre Expert
T +32 2 339 79 62
[email protected]