This tip is part two of a series on using data center infrastructure management tools. Read part one: data center management using DCIM tools.
The DCIM market includes established IT players, such as IBM, Hewlett-Packard Co. and CA, and relative start-ups and newcomers, such as PowerAssure Inc., Viridity Software Inc. and Modius Inc. Major infrastructure vendors have been trending toward consolidating themselves to be a one stop shop for all data centers needs for quite some time. Emerson Electric’s acquisition of Avocent Corp. last December and Schneider Electric’s purchase of APC in 2007 are two examples.
Other well-established, but somewhat smaller firms, such as Raritan Inc., have expanded their IT offerings to focus on the functionality of iPDUs with KVM solutions in a central management console. IT administrators can open remote KVM sessions with servers, monitor the power used by each individual device and turn on and off the power to each receptacle on the iPDU. iPDUs can also track each server’s power usage to help monitor and improve the computing load versus the power used.
Gartner Inc. projected that the data center infrastructure management (DCIM) market will experience very high growth in the next few years. The lure of an expanding DCIM market even has vendors such as Panduit Corp., which traditionally manufactures network cabling systems and rack-related hardware, to release a DCIM offering.
Ultimately, the most likely winner in the DCIM market will be companies that have packages for interfacing and interchanging data that is meaningful to IT and facilities teams. To put things into perspective, the IT department is concerned with upgrading the network cabling to CAT-6a, for example, to meet the rising bandwidth requirements, while the facilities department is concerned whether the six-inch pipes carrying the chilled water need to be upsized to meet the rising heat load.
Associated costs of DCIM tools
The total cost of a DCIM implementation project can be broken down into three key areas: the price of the DCIM platform and software (base cost and per node licenses), the cost of the energy-monitoring hardware and environmental sensors, the installation cost of the required hardware and wiring, as well as any interfacing and systems integration expenses for existing hardware or systems.
When considering various DCIM vendor platforms, the ability to integrate and accept data from existing systems is a key evaluation point. Moreover, the number of devices (or managed assets) to be monitored has a direct relationship to the final overall cost.
The higher the level of granularity (or the more devices that you need monitored) the more costly and complex the DCIM installation becomes. It is very expensive to install more hardware to monitor the energy usage at key infrastructure points -- chillers, cooling towers, individual CRACs or CRAHs -- as well as lighting and generator panels. Moreover, if per rack monitoring is implemented, the iPDUs could add substantial costs to larger sites. In addition, most DCIM systems’ licensing costs are based on the number of monitored devices, and you also must factor in the price of the system’s software. A basic project can cost approximately $10,000, while a large-scale, highly-granular deployment can run as much as six or seven figures.
The bottom line
The DCIM category is a result of the somewhat reluctant acceptance of the need to measure and improve energy efficiency in the data center. The goal of DCIM is to help both IT and facilities managers work together to make more informed energy optimization and capacity planning decisions.
The total integration of IT architecture and facilities’ energy-management systems into an energy-aware computing architecture is the holly grail of data center efficiency. Actually automating and optimizing energy and computing efficiency in real-time, rather than just providing information for management review, is the ultimate goal. If this goal is fully realized, the result would be a truly dynamic data center. The relationship between the actual computing load and the total energy used would be in tune, unlike in the present data center, where the power is constantly consumed at a high rate, regardless of whether IT systems are idle or running at full load.
Back to part one: Data center management using DCIM tools
About the author
Julius Neudorfer is the CTO and founding principal of NAAT. Neudorfer has designed and managed communications and data systems projects for both commercial clients and government customers.