The cloud is expanding beyond the confines of the traditional server farm.
In our last article, we touched upon how the growth of internet-connected devices is fuelling massive data centre growth in APAC and worldwide. Logically, such trends beg the question: How will data centre infrastructure management, or DCIM, have to adapt in response?
First, let’s look at some examples. With growth on pace for $4.5 billion by 2018 and recognition from Gartner as the second-fastest growing data centre market in APAC, India is becoming even more of a hotbed of activity. In light of such expansion, the nation’s ongoing transition from outsourced data to cloud computing could be a key indicator of where the rest of the world might be heading.
Data centres need to cope with the fact that even as usage migrates towards Infrastructure as a Service, or IaaS, the cloud is expanding beyond the confines of the traditional server farm. MarketWatch reported that the drone data service industry will possibly be worth $8.15 billion as soon as 2025. As more companies use unmanned aerial vehicles to capture imagery, the cloud must be capable of the analysis required to make sense of it, and this means data centre deployments will require increased flexibility and heightened connectivity.
Data services (IOT) are also revolutionising formerly untouched fields. Caterpillar, for instance, sells subscriptions to sensor data that it collects from some 500,000 of its devices. Waste management companies are partnering with data-as-a-service providers to master information processing and put captured business data to good use. Data centre operation is no longer solely the purview of enterprises traditionally thought of as IT companies, and as the Internet of things (IoT) swells, practices like DCIM will also grow in importance.
As you may expect, uncertain legal frameworks are also having an impact. Even as national law enforcement bodies push for greater oversight rights, agencies like the U.S. FCC are taking steps towards deregulation. For instance, an April 2017 Report & Order by the agency eliminated price caps on packet-switched business data services. While not strictly in the domain of data centre operators, such changes could impact costs of provisioning bandwidth to keep their clients connected. With data usage proliferating at such overwhelming rates, data service providers need to assess ways to control their overhead so that they can continue offering competitive products.
It’s also worth considering that growth isn’t always smooth. Quick jumps are common, such as when Indonesia-based telecom provider XL Axiata saw a 218 percent increase in data traffic during the 2017 Eid al-Fitr holiday. Consumer-oriented companies whose service offerings revolve around information must ready their infrastructures for unexpected stresses. Those who don’t are almost certain to suffer consequences for their unpreparedness.
Just as garbage collectors, drone pilots and heavy equipment operators are using Big Data to improve their operations, data centre administrators need to hone their oversight techniques. Moving forward, intelligent DCIM systems that deliver accurate feedback will become essential to improving DC efficiency, reliability, and predictive and proactive maintenance.
To deliver on DCIM, some are choosing a tightly-integrated proprietary DCIM and DC hardware such as Baselayer. Others propose adopting a common open model such as Open19 to make the DC more efficient through standardised rack, server, and shared power, and one could only assume eventual integration with DCIM. But more frequently we see traditional DCIM systems integrated – with varying levels – over different vendors.
What are your thoughts?