software & applications
CLEAN AND CORRECT
Data centre analytics solutions are only as good as the data they use . However , sensors and collection methods are prone to errors , can produce bad data or generate no data , and generally can be unreliable . Richard Jenkins , SVP of global marketing and strategic partnerships at Romonet , explains why it ’ s important , even before you invest in management software , to make sure your data is validated , certified and clean .
Despite analytics being one of the main buzzwords of the moment , organisations are still unsure how to utilise solutions . There is a vast pool of data available on every aspect and function of a business , but data needs to be analysed accurately if it ’ s to be useful .
Data is being used across the business to improve processes and performance . Every department , from marketing to finance , is considering how analytics can be used as a beneficial tool to provide a deeper level of understanding of business processes .
In the rush to keep ahead of the crowd , businesses are investing in projects without considering what they want to achieve or how to gain value from the information gathered .
Looking within
The data centre , the home of an organisation ’ s data , is ironically one of the last areas to be considered for an analytics project .
The increase in data as a result of analytics and Big Data projects will typically affect the performance of the entire data centre estate – from owneroperated to cloud and colocation .
For many organisations , the ability to provide IT services to the edge will be the difference between a fast growing business and a steadily growing business . To handle this level of data , the data centre estate needs to be managed and analysed in detail .
Therefore , any reports on a facility ’ s financials , operational performance and environmental profile – cloud , colocation or onpremise – need to be verified as accurate .
‘ Bad data ’ after all leads to bad decision making . Executives assume that reports produced by operational teams are correct but
26 | March 2017