Lara Greden is a senior principal, strategy, at CA Technologies. Her previous post was titled, Using DCIM to Achieve Simplicity in the Face of Complexity. You can follow her on Twitter at @laragreden.
DCIM technology is core to Facebook’s approach to managing capacity and achieving efficiency in the data center. At the recent Open Compute Summit in San Jose, California, Tom Furlong (here's a video link to his talk) spoke about how Facebook defined and implemented DCIM technology to accommodate 42 use cases.
As part of their process, Facebook conducted a Proof of Concept (POC) with third-party DCIM vendors to collect and rationalize thousands of data points in real-time, or near real-time, before investing in a platform.
While Facebook may have different criteria in how they approach DCIM, you may want to consider the following factors before deciding how to bring DCIM into your data center:
Ensure that the vendor’s DCIM technology has the scalability and architecture to meet the specific design of your data center or portfolio of data centers, whether owned or leased from a colo provider. You will be pulling on the order of hundreds or thousands of data points. Start with an evaluation of your requirements, and then task the vendors to demonstrate how they can achieve those requirements in your environment. Through product demonstrations, interviews, site visits or, if appropriate, a proof of concept, you want to make sure that the scalability and architecture will produce the results you require.
Know It Will Integrate
One of your requirements is likely something along the lines of “gather data points from across the physical equipment and systems in our data center, including CRACs/CRAHs, variable speed fans, chillers, BMSs, branch circuit monitoring, PDUs, UPSs, generators, servers, sensors, and more.” Your data center uses equipment from a variety of vendors, so you want to be sure the solution will provide integration to meet your use cases, both now and in the future.
One of the aspects to consider is how the solution integrates with the physical data center. Evaluate that the data gateway supports the varies device protocols and frameworks necessary, such as SNMP, XML, CSV, IPV6, BACnet, Modbus TCP/RTU, WMI, IPMP, EnergyWise and others. Also assess the requirements, if any, for installing, configuring, and maintaining additional hardware.
Another of your requirements is likely to be crafted around the need to integrate upstream with IT Management systems, such as the service desk, workload automation, change management, incident management, and native workflow solutions. Ask for concrete examples of how the DCIM solution will integrate with the systems, processes, and workflows you’ve defined in your requirements study.
Refine Your Requirements Study
To ensure successful adoption of DCIM technology, your program design should be grounded in a deep understanding of the end users and end goals. As a result, you should approach DCIM from a program perspective instead of a project perspective, and understand which requirements will put you on the road to success. As with any large scale project, start with the use cases and requirements that will provide immediate value and build from there. Think of your requirements as a roadmap, and use the early learnings from the first phase to help refine requirements for subsequent phases.
One of the first areas of value realization from DCIM comes when managing data center systems manually is no longer cost effective. By bringing together the data sets with DCIM, you will realize greater value due to visualization of combined data sets and intelligence based on analytics across the data sets. DCIM often comes to the forefront when data center consolidation programs are underway, and it is essential when data center capacity management is a core functional discipline in your organization.
What Are Your Objectives?
As with any enabling technology, there is no single reason for deploying DCIM technology. For some, it is about capacity management. For others, it is about uptime and availability. For others still, it is about efficiency. But most likely, it is a combination of all of those mission-critical objectives. As your organization goes through the process of evaluating DCIM technology for your data centers, build on the wealth of lessons learned in the industry and give particular scrutiny to how DCIM can be fine-tuned to meet your business objectives.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.