OPINION
It all begins with the data
Regardless of the chosen strategy for their facilities, what
most data centre operators don’t realise is that in order
to increase their facilities’ performance, firstly, they must
clearly assess and understand their current state. Only
when they know the exact areas that aren’t as efficient
as required can they truly make significant changes and
have the certainty they can deliver the expected business
outcomes. The challenge of achieving this level of clarity
and reliable insights into data centres’ health and potential
performance opportunities stems from the poor quality tools
and data many companies are still using to operationally
and financially manage these facilities.
As the IoT is evolving and the number of sensors and data
collection points is skyrocketing the situation is bound to
become even more challenging for data centre operators.
It’s already humanly impossible to manually verify and
clean the vast amount of data constantly streamed from
these sensors. To counteract this, an increasing number of
organisations are now adopting smart analytics solutions
to improve the forecasting, construction, management and
analysis of these facilities.
Clean data for smart analytics
Many data centre managers using DCIM tools don’t realise
that in many cases, even if a meter or sensor is highly
accurate, the data streamed and collected from it can’t
be used in its raw state. In reality, experience shows that
on average only 60-65% of raw data is even suitable for
cleaning, validation, normalisation and labelling prior to
being used for any sort of analysis. Therefore, while analytics
solutions and DCIM tools are great for improving the
facilities’ environmental impact, reducing energy costs and
maximising performance, data centre operators need to also
ensure the quality, accuracy and ultimately the reliability of
the data they collect for analysis.
By using data cleaning tools and techniques, data centre
organisations can be sure that the data they base their
strategic decisions on is accurate and validated. Considering
how expensive data centres can be, making the right
decision about the region in which to build the data centre,
understanding the climate’s impact on construction and
maintenance and forecasting what equipment or cooling
system is needed for the facility to offer reliable, high quality
services long term but also scale up and down according
to business priorities can make a huge difference in a
company’s success or failure.
Data centre operators that want to maximise
performance, reduce the costs of energy consumption,
increase profitability and decrease facilities’ environmental
impact should begin by cleaning their data and making sure
it’s accurate and reliable. Then they can truly understand
their facilities inside out and clearly assess what areas are
performing and what areas need improving.
Other critical questions data centre stakeholders need
answers to are: What are the main risks facing the facility
in the short and long term? Is the equipment calibrated
for maximum performance? What opportunities are there
for improving energy efficiency? What needs to be done in
order to adjust the data centre’s capacity? The answers to
these questions can only be found by using the right tools. n
www.networkseuropemagazine.com
41