Part Two: The Converged Infrastructure and Your Power Bill

In the first part of our discussion on the “converged infrastructure,” we detailed why we agreed with HP that the converged infrastructure represents the data center of the future where all your IT resources and systems exist and function dynamically.

We discussed how this creates a greater need for energy-efficient IT solutions capable of supporting required increases in computing performance and density while reducing overall data center power demand and lowering operating costs.

This brings us to a new point: what is the best way to view the power and cooling impact of the converged infrastructure?

In the HP article on converged infrastructure that originally gained our interest, they say that “a data center is by nature a highly complex system filled with IT hardware and racks, connected by miles of wires and cables, and complex relationships between hardware and software.”

Yes, that sounds like most data centers to us.

Then they go on to say, “IT intersects with another complex system, the data center facility itself, delivering power, cooling and space for IT. For reliability and availability, both IT and the facility are usually over-provisioned. An additional and fixed energy and cooling buffer is usually allotted to make sure that critical peaks of IT usage do not result in IT meltdowns. Plus, power and cooling is “fixed,” usually supplying a constant amount to IT—no matter the data center status or workload.” We think that this “intersection” is critical, since it occurs at the enclosure level. So your enclosures shouldn’t be thought of as a commodity. They are an integral part of the whole system.

This is where we think HP has really captured the essence of the challenge. Most data center facilities are over-provisioned and under-utilized – until they try to support increasing computing densities. While the overall data center may be provisioned to handle the higher densities, hot-spots can develop in individual enclosures. So an enclosure that is smart enough to utilize the available cooling is critical.

We use the “Four Ms for Datacenter Energy Efficiency” as a framework for thinking about how data center professionals can manage power and cooling more efficiently. Understanding the mechanisms of these different components will help ensure greater control.

Measure: For increased efficiency, the supply, server inlet and return must all be measured. Airflow must be measured along with power consumption and air temperature. A lot of cooling is wasted, which is not measured otherwise.

Manage: Airflow management is a key factor in data center efficiency. The flow of supply air should be managed as closely as possible to the computer inlet. Equipment such as variable speed fans at the air-handler or at the rack level, and high-efficiency chilled water pumps help you maintain airflow at the cabinet level.

Monitor: Cabinet level is no longer an option; it is a mandate, especially with increased densities and power/heat loads. This process eases managing the data center against ‘downing’ events.

Maintain: Use of the most successful ‘best practices’ is the best way to keep the datacenter environment efficient and in tune with technology turnover. Systems should be sized and redundancies configured to maximize efficiency, while air-handler systems should be rebalanced.

We focus on using proactive performance tools that utilize a DCIM platform, along with intelligent enclosures to collect, store, share, analyze and report information on power, space and cooling – at the enclosure level, in real-time – across the data center.

What do you think?

Part One: The Converged Infrastructure and You

HP has been talking quite a bit about a “converged infrastructure,” which our trusty Wikipedia defines as packaging:

“Multiple information technology (IT) components into a single, optimized computing solution. Components of a converged infrastructure solution include servers, data storage devices, networking equipment and software for IT infrastructure management, automation and orchestration.”

We think this means that the converged infrastructure represents the data center of the future where all your IT resources and systems exist and function dynamically.

We’re particularly interested in the way HP maps out the effect of power and cooling in a converged infrastructure. According to HP, a converged infrastructure “creates an intelligent,
energy-aware environment across IT and facilities to optimize and adapt energy use, to reclaim facility capacity, and to reduce energy costs. It collects and communicates thousands of power and cooling measurements across IT systems and facilities in real time to give your organization greater insight and control over energy use.”

Let’s take a look at what they’re saying.

We like the idea of an energy-aware environment that encompasses both IT and facilities. We’ve long advocated that these disciplines be cross-trained and that organizations need to bridge the IT-facilities gap. But we think HP gets to the real heart of today’s data center requirements when it discusses the need to “collect and communicate thousands of power and cooling measurements across IT systems and facilities in real time.”

We mentioned a few months ago that Gartner analyst Ray Paquet had noted the “Top Trends to Watch for the Data Center.” Among these were Energy Efficiency and Monitoring and Converged Fabrics. We’re thinking that the HP Converged Infrastructure and AFCO’s active enclosures align nicely here.

Bottom line: There is a great need for energy-efficient IT solutions - including servers, storage, networking and communications systems, as well as energy-efficient cooling techniques - that support increases in computing performance and density while reducing overall data center power demand. Managing these often competing resources correctly can save a large part of the IT infrastructure bill and lead to lower operating costs.

What do you think?