Page 253 -

The Greening of IT
218 How Companies Can Make a Difference for the Environment

6. Consolidate on large servers. Fewer larger servers will convert AC
to DC more efficiently than many smaller servers with smaller and
less-efficient power systems. Power supplies of large servers are
capable of operating at very high efficiencies (+90 percent). Large
servers can also take advantage of high voltages and eliminate a con-
version loss that robs efficiencies when stepping down to smaller
voltages. The ability to more efficiently share resources makes run-
ning on a few larger systems more efficient than many small ones.
Workload can be balanced, driving up utilization and reducing the
number of watts needed to run applications, day or night.
7. Eliminate redundancy but keep high availability and disaster
recovery capabilities. High availability and disaster recovery can
be efficient and in a green way be designed into server configura-
tions. Engines can now add nondisruptively to almost all platforms,
reducing the need for extra servers. No longer is an idle server
needed for what-if scenarios. Production servers can back up other
production servers. Configuring the ability to nondisruptively add
(and reduce) capacity for production or disaster recovery without
ptg
having idle or underutilized servers significantly reduces the num-
ber of footprints and slashes the energy consumed in the data center.
Commonly used technologies include IBM’s On/Off Capacity on
Demand (add engines by the day) and Capacity Backup Upgrade
(CBU) for Disaster Recovery. A data center can be greened and the
bottom line affected by using fewer servers while having the ability
to increase capacity without adding server and the associated facility
infrastructure.
8. Measure and put the costs of energy where they are incurred.
Automated measuring and billing of energy consumption makes
usage part of cost and green decisions. Without energy and cooling
knowledge, requirements are unknown, inaccurate, and often over-
planned leading to inefficiencies. An example of new technology to
optimize energy use is IBM’s Active Energy Manager (AEM).
Monitoring energy usage and developing trends is key to under-
standing how energy is being used. This first step to optimizing
energy use opens up the potential to become more efficient and
optimizing for performance / watt. Managing energy use is an evolv-
ing concept in the data center. Capping power at the server level
and optimizing to deliver the right performance per watts can be
achieved using AEM. In the future, the most efficient data centers