12 / 88
12
7X24 MAGAZINE FALL 2015
Data center infrastructure management—real,
dedicated DCIM—is about a decade old (barely),
and already has evolved into something its creators
didn’t expect. DCIM was designed for a different
time and a vastly different data center—the swollen
facilities of the early 2000s that housed row after
row of servers and were built to add even more. No
one could imagine a reason why things wouldn’t
continue that way, and organizations needed to
track and manage those assets in increasingly
virtualized environments. Because necessity, as
they say, is the mother of invention, DCIM was
born. No, scratch that. This was when DCIM was
conceived. Birth came later, and that’s an important
distinction.
DCIM gestated while the need for greater visibility
and asset management increased. But something
else happened during this time—the data center
evolved. Server capacity increased, energy
efficiency across systems became a priority and
enterprise data centers became something of an
endangered species. Hyperscale—think Google,
Amazon and Facebook—emerged with its own
unique architectures, and cloud and colocation
facilities started to siphon computing away from
traditional enterprise facilities. As DCIM became
fully formed, it emerged as and remains a valuable
solution in these environments.
But there’s another vein of data center evolution—
distributed networks, with IT closer to end users
and smaller data centers at the hub. These
networks can have multiple small computing
modules spread across different locations,
providing local computing and storage while still
networking with each other and with the small data
center at the network’s center. Individually, these
are simple IT resources—often just a rack with a
server or servers, power distribution, maybe some
by
enzo Greco
Cambridge University simplifies IT operations through data center infrastructure management