by Mike Kaul

Data Center Infrastructure Management -- or DCIM -- is on everyone’s minds these days, and for good reason. According to Forrester Research, DCIM is expected to grow to 60 percent market penetration by 2014, up from just one percent market penetration in 2010. While I’m often skeptical about industry prognostications, the major trends out there that will fuel DCIM adoption are undeniable:

• Data center consolidation is accelerating;
  
• Enterprises are moving to virtualization and private cloud computing;
  
• Businesses are increasingly reliant on their critical IT systems;
  
• Data Center capacity limitations are impacting service delivery growth;
  
• Increased power consumption and cost is a major problem.

Of course, when such market growth projections are made, many companies rush to enter the market. This was true of DCIM in both 2010 and 2011. In fact, I know of more than 30 companies that count themselves as DCIM vendors today -- and there are probably a few dozen more that simply haven’t crossed my radar yet.

The problem with all the noise in the market is that it becomes difficult for customers to figure out which solutions will meet their current and future needs. They know they can’t ignore DCIM, but how do they figure out which “big rocks” they need to move? (Let me give you a hint: It’s not by simply tracking heating and cooling.)

Now, lest anyone misunderstand that last statement, I think it’s a good idea to track cooling and temperature and there are solutions that track such things very effectively. That said, to solve the big problems that CIOs have, the focus needs to be on the IT part of the data center. Why? Because those are the hills where the gold is.

A CIO can take an enterprise data center and spend about $100,000 to optimize the cooling, HVAC, airflows and overall environmentals. Yet, even after doing this, that CIO will still have servers running at about 12 percent capacity and VMs running at 7 to 8 percent capacity. He or she will have no idea which applications should be virtualized, moved to the private cloud offering being considered, or exactly how much energy the company’s storage infrastructure is consuming.

Yes, the real return for CIOs comes with effective capacity planning. With continually constrained budgets and ever more services to deliver, cash back is not what today’s CIOs need. What they need is additional compute cycles, more I/O to the discs, and heavier traffic on their networks. To make this happen, CIOs must do a number of things.

1. They must gain granular insight into the entire IT stack and key metrics such as compute cycles per kilowatt hour, terabyte stored per watt, etc.
   
2. They must also find their stranded capacity, get it back, and put it to work for the business.
   
3. Then, they must slow down capex on new equipment until their existing equipment is operating at well over 50 percent and all VMs are at 90 percent.
   
4. They should also look at their top five applications and understand exactly how moving them to that private cloud or virtual cluster will deliver increased capacity and efficiency. They must really understand their current level of efficiency and utilization, right down to the bare metal and service/application.
   
5. Finally, once they have all this insight, they must find a way to analyze their options and model the most effective ways of operating their data center infrastructure going forward.

If this sounds to you like what Gartner termed Intelligent Capacity Planning -- or ICP (DCIM: Going Beyond IT, Gartner Research, March 2010), then you’re right. ICP makes possible the aggregation and correlation of real-time data from heterogeneous infrastructures, giving data center managers a common repository of information on resource utilization and performance. According to that report, “This functionality will become a key component of IT organizations' capabilities, if the optimal capacity, energy efficiency and computing potential of a data center are to be achieved.

Through ICP, data center managers can optimize the performance, reliability and efficiency of the entire data center infrastructure. This is achieved via automated management of IT applications based not only on server capacity, but also on the conditions within the physical infrastructure of the data center itself.

For any CIO contemplating virtualization and cloud technologies, the synchronized monitoring, predictive analytics and management of both physical and virtual infrastructure in a highly dynamic environment is going to mean the difference between the business viewing IT as a cost center and the business viewing IT as a strategic business weapon.

I truly believe that with the rapid continued growth of business-critical IT applications and services, intelligent capacity planning will be the yardstick by which CIOs will be measured. And that Data Center Analytics (DCA) will be at the very heart of intelligent capacity planning.

Why?

Because choosing the right path every time can only happen if you are able to quickly and easily simulate and analyze all of your viable options.

Mike Kaul is the CEO of Sentilla Corporation (Redwood City, CA). www.sentilla.com