The networking environment has changed drastically over
the last decade and there seems to be no slowing down as consumption of data
begins to move from static workstations to mobile devices with access
everywhere at any time. This transformation of how people are doing business
and are leveraging mobile computing devices such as phones and tablets is more
than a game changer, it is a life changer.
This change is not only happening for the individual but
also at the other end of the system, the Data Center. Mobile computing has its
own challenges and issues, such as access, scalability and speed. As more and
more information is used when and where it is required these issues are
becoming a greater concern. Connection to the cloud (aka Data Centers) is the
key to service. According to http://www.statista.com/statistics/271405/global-mobile-data-traffic-forecast/
global mobile traffic is almost doubling per year, and the amount of Exabytes
of information being shared per month will increase by 10X in the next 5 years
from 2.5 to 24.3. That’s a lot of data.
Not only is data steadily increasing, but we have already
experienced sporting and social seismic events which can impact systems
radically in a short span of time. For example during the last Superbowl, 4.1
Terabytes of data were generated compared to the previous years of 1.7
Terabytes, according to http://www.fiercewireless.com/story/verizon-att-sprint-and-t-mobile-reveal-network-traffic-stats-super-bowl/2015-02-04.
These spikes in usage require data systems to be flexible and scalable enough
to handle “on demand” traffic. Capacity from a user perspective is now considered
a given no matter where they are; expectations are that service should be
everywhere and available at all times.
In order to keep devices as user friendly and portable as
possible, software is no longer being housed at its access point. Apps and
virtual machines have taken over the bulk of processing within the network
landscape and the SaaS model has allowed mobile devices to become flexible
platforms and interfaces that can do almost anything. Enter the modern day Data
Center, the brains behind this technological brawn. They are not only a
facility to store, house and process information; but also a platform that
needs to maintain scalability to deal with the influx of users accessing a given
program at a single point of time.
The Data Centers main function in this scenario is to
support the networking and data environment, while at the same time providing a
suitable environment for the physical infrastructure. And since the nature of how
we access and use data is changing, so to must our Data Centers. Data Centers
are very complex environments that have many different components at both a
facilities level as well as a networking level. One system impacts another,
which makes scalability difficult to achieve. To sit down and draw out how a
singular change in equipment sends ripples through the DC environment, would be
like trying to count the ripples in a lake once you threw a boulder into the
centre.
Each piece of hardware is unique in its power
consumption, heat generation and cooling requirements and although you can
guestimate its effect on an environment using the operational specification
ranges, until you turn the unit on in a given environment you won’t know its
impact. The closest you will be able to come to a projecting a Data Centers
power, cooling and capacity, is in a virtual environment.
Virtual
environments allow you to build a data center without actually building a data
center. You can run comparisons of different equipment, heat profiles,
efficiency profiles, rack densities and even projected growth models that would
take care of scalability. Making a Data Center as efficient as possible is the
key to controlling the costs within a Data Center. Data Infrastructure
Management (DCIMs) not only allow
for extensive and comprehensive planning but also comparison of projected and
real environmental offsets.
In my opinion this is the only way anyone should be
building Data Centers today. The granularity of facility data produced through
the monitoring of a live data center allows the user to make intelligent
decisions about power and cooling costs as well as tracking projected hardware
change impacts. In essence we are actually comparing Data Centers without DCIM
and those with DCIM. I am a numbers guy and that is the way I make personal
financial decisions, and would also make business decisions. Without monitoring
where money is being spent and how much on what, there is no hope of realizing
the full potential of cost savings which can be achieved by fine tuning and
optimizing the Data Center.
Just
like we have SMART phones, we need SMART Data Centers to
support them and that means DCIM.
For more information on DCIM products call
1-800-427-2055
No comments:
Post a Comment
Note: only a member of this blog may post a comment.