This is the first part of our five-part series on the countless number of decisions an organization needs to make as it embarks on the DCIM purchase, implementation, and operation journey. Read Part 2 here.
The advent of DCIM as a term originated not long after The Green Grid introduced its now widely recognized Power Usage Effectiveness “PUE” metric in 2007. A substantial percentage of early DCIM offerings were primarily marketed as PUE dashboards in order to gain better acceptance and recognition of the drive to measure and improve energy efficiency in the data center facility.
DCIM has evolved to become a very broad term, widely utilized by a myriad of vendors that have depicted it as the ultimate data center management tool. So is DCIM mostly market-driven hype, or are there tangible benefits to be realized? Considering the costs of products and implementation, where is the return on investment? This five-part series will exam the business benefits, pain points and processes that DCIM product suites are attempting to deliver or improve.
I have written about DCIM previously but have noticed a significant uptick in interest this past year by potential end-user organizations - not just increased marketing efforts by the major players in the data center market. However, the data center landscape itself has undergone significant changes since DCIM was first introduced, and many enterprise organizations have moved toward colocation and cloud services to meet increased capacity demands, rather than build more data centers. In response, some colocation providers are implementing DCIM as a selling point to their clients by offering them a peek behind the curtain, and providing their customers access to selected aspects of their internal DCIM systems.
Originally, some of the early DCIM products were based on adaptations of traditional Building Management Systems (BMS) platforms. They were almost totally oriented toward energy usage, so that the facilities' team could monitor, manage and hopefully optimize the power consumed by the electrical and cooling systems. And like the PUE metric, these were facility-centric packages that did not correlate with the energy efficiency of the IT hardware (or any other aspects of the computing systems). In contrast, IT administrators have long had a wide variety of administrative management consoles and tools that look at different aspects of their servers, storage and networks systems. However, in many cases they were also somewhat segregated and specialized, especially in larger enterprise systems; and virtually none of these systems examined IT energy usage or efficiency.
The long-term promise of DCIM is to help both IT and facilities managers work together to make more informed, and presumably better, decisions about overall energy usage - operational and workflow optimization - as well as capacity modeling and planning. As the name implies, it is rational to infer that DCIM presumably provides organizations with the ability to monitor and manage their infrastructure. However, if you ask what “infrastructure” means to different factions of the data center world, you will still get diverse answers. The same seems to hold true for today's DCIM offerings by various vendors.
The current DCIM options cover an extensive array of products with a wide range of features and functions. Conceptually, they are based on a centralized data repository that can deliver an integrated and interrelated view of all the assets and status of the physical facility infrastructure (space, power, cooling and network cabling) as well as IT systems (servers, storage, networking and even applications). Nonetheless, in many organizations facilities and IT teams are traditionally siloed: culturally, technically and politically.
The Politics of DCIM
Given the historically divergent cultures of IT and facilities, it should come as no surprise that politics play a significant role in DCIM decisions. This tends to apply more to traditional data centers than to new build-outs. It also seems to relate to the type of organization: conservative financial firms vs. internet-based services, such as cloud, search and social media.
The new economic reality focused on cost reduction has resulted in the rising costs of energy becoming a significant percentage of the OPEX. Suddenly, PUE became a buzzword that even the CFO heard about; and at least awareness of the need for basic energy efficiency measurements become a Key Public Indictaor (KPI). However, justifying and maximizing the promised DCIM value proposition and the buyer's motivation can be at opposite ends of the spectrum. Clearly, senior management wants better reporting metrics and lower OPEX. However, the technical staff - which should be directly involved with the product evaluation, testing, and purchasing recommendations - may have mixed motivations.
While DCIM ultimately should make IT departments more productive by improving and automating processes, like any labor saving system it could eventually reduce head count. Prior to the 2008 financial crisis, data center managers were primarily motivated to provide security and 24x7 availability for their facilities (the proverbial five “9”s), and to satisfy customers' requirements.
Employee productivity was important, but maintaining qualified personnel that kept everything operational was even more so. In effect, as long as power and cooling met IT requirements, there was very little reason for elaborate facility efficient metrics. While general operating costs were effectively kept in check (in-house staff, vendor-provided equipment services, etc.), availability became the KPI. In essence, this binary indicator assigned a facility a 1.00 if it had no outages or a 0.00 if there were any (in which case fingers are pointed and perhaps heads will roll).
The politics of social responsibility, carbon and water usage metrics (CUE and WUE), and disclosure also come into play. Prior to the development of the Green Grid metrics, very few organizations really tracked data center energy efficiency (much less water usage and carbon footprint). There was not much to report either internally or publicly, and data centers never attracted the attention of watchdog organizations such as Green Peace. Now it seems that even those organizations that strive to build new facilities with PUEs below 1.2 or even 1.1 are still targets and subject to criticism - despite the vastly improved sustainability performance of new facilities.
One unclear issue for facilities operators is the difference between existing Building Management Systems (BMS) and DCIM, and any additional benefits it offers. IT teams are also very reluctant to interconnect their systems to the facilities systems for a variety of reasons. Most recently, security has become a major concern and impediment.
In speaking to a variety of DCIM vendors, several customer-driven issues have emerged. They have seen that some of the end-users expect to automate or manage workflow processes that may not even presently exist as well-defined manual processes (i.e. when IT servers need to be added, operations and facilities should effectively allocate and provision space, power, cooling and network, etc.). This in itself indicates a basic visibility gap that needs to be addressed, and DCIM may help with organization and management.
So, how can CEO-level management executives interpret and evaluate the usefulness of the promised deliverables, much less who should specify, purchase, install and operate DCIM systems? Clearly before embarking on a trip down the DCIM wormhole, a team composed of facilities and IT needs to agree and address the major pain points and primary areas of improvement by taking a holistic approach to energy efficiency, workflow process, and asset management, as well as capacity optimization, modeling and planning.
The Bottom Line
DCIM has now moved far beyond a glorified PUE console as it embarks on its second and third generations as a product category. However, despite the early hype and projections of massive and widespread adoption by 2015, DCIM is still a complex solution for a multidimensional problem. Plus, it's being promoted to a somewhat skeptical and perhaps confused audience of potential buyers.
What are the drivers and key motivators for data center decision makers to implement or defer DCIM deployment? Should DCIM be viewed as a strategic investment or just another additional (perhaps unnecessary) expense? We will address these issues over the course of this series, examining details behind the decision to buy or not to buy DCIM, its benefits, implementation challenges, expenses (direct and hidden), and ultimately the justification of cost.
This was the first part of a five-part series on DCIM. Read Part 2, which talks about considerations necessary before purchasing a DCIM solution, here.