Untitled Document
Editor's note: Missed our SOA in Action Virtual Conference? Replay
sessions on demand here.
Cloud computing promises a lot: access to large-scale virtualized computing
and storage, 24/7 availability and costs associated solely with the resources
used.
The cloud is recognized to have started in 2006 with Amazon Web Service's (AWS)
introduction of the Elastic Compute Cloud (Amazon EC2). Incredibly innovative,
this service, when combined with the Simple Storage Service (Amazon S3), enabled
developers to start virtual machine images and consume storage as a service,
where billing was done on per-hour and per-gigabyte rates.
These services were the start of what is now recognized as Infrastructure-as-a-Service
(IaaS). Currently, managed hosting companies, including GoGrid and Rackspace,
are integrating IaaS capabilities to provision external resources dynamically.
It's expected that more providers will follow Amazon's lead.
Efficiency challenges for data centers
In contrast, traditional internal data centers have been grappling with record-high
energy prices, fueling the demand for efficiency within an organization's infrastructure.
According to statistics from the Uptime Institute, a 451 Group company, the
average data center is 10 percent utilized, and therefore 90 percent idle.
Additionally, only about 3 percent of a data center's electricity consumption
is spent crunching data. The remainder is spent on idleness, redundancy and
other energy overhead. Data center electricity use is growing at about 12 percent
per year, which creates the need to add about 1,000 megawatts of electricity-production
capacity annually. This is equal to building two large 500-megawatt power plants
at a cost of $1-$2 billion each construct.
All of these costs and inefficiencies could be saved by increasing utilization,
saving energy and, in the process, saving money. Prior to the advent of virtualization
and the cloud, this inefficiency was a requisite for decoupling applications
from each other and enabling reliability through redundancy. Critical applications
had a production and disaster-recovery environment, both fully provisioned in
separate locations to ensure that uptime could be guaranteed even with outages
and failure.
-1-