IO Blog

Going forward, our blog will feature posts from IO’s global regions, with content  delivered from the local different perspective. This piece is from IO EMEA:

The problem of data centre sprawl—increasingly complex and granular environments and the new demands on infrastructure by the software and hardware it supports– is typically addressed by DCIM (Data Centre Infrastructure Management).  It’s served the industry well, but it’s starting to reach its natural limits, and we think it’s time to build on the foundations DCIM laid down.

“DCIM tools provide insights and improve performance throughout the data centre, including data centre assets and physical infrastructure,” said Gartner analyst David Cappuccio in a report on DCIM published in 2012. “They enable the monitoring and collection of low-level infrastructure data to enable intelligent analysis.”

DCIM is fine for understanding and asserting basic control over a distributed, varied estate that predates the DCIM concept, but more and more data centres were built after the advent of DCIM – and have surpassed technologically what DCIM was intended to address. What’s needed is software that unifies all of the layers in a data centre – what we call a Data Centre Operating System (DCOS).

The need to unify distributed data centre assets is one reason for a DCOS, but there are others; the data centre has become the computer, with applications sitting somewhere in it – rather than on specific pieces of hardware. What might have been clusters in the past are now serious computing platforms. All of this needs a coherent management tool to get the most out of it. A DCOS provides that single platform, but also enables resource sharing, data sharing and programming interfaces for developers.