The shifting of more heavy-duty production workloads into virtualized environments is transforming the data center (DC) and the upcoming wave of the Internet of Things looks set to create more pain points for infrastructure managers.
No surprise then that annual spending on data center infrastructure management (DCIM) software and services could grow to more than USD4.5 billion worldwide by 2020 from USD663 million in 2013, driven by DC operators’ need to improve energy efficiency while meeting ever-growing demand for IT capacity, according to Navigant Research.
These challenges, Gartner analysts observe, have fuelled growing interest in DCIM beyond power and cooling management to capacity planning and predictive analysis as operators also seek optimal use of physical space and assets.
Organizations Gartner analysts spoke to that either had already deployed DCIM or were planning to by end 2015 commonly used capacity planning as the primary justification for the purchase. This was followed by predictive analytics, with power and cooling management coming in a distant third.
Another study by 451 Research analyst Rhonda Ascierto estimates that power and environmental monitoring, together with asset, change and configuration management, account for about 80% of the total DCIM market in revenue terms.
Already, the need to control rack level energy consumption and cut down operational costs are compelling large and mega DC operators in Asia Pacific to look beyond the contribution of power and cooling equipment to comprehensive DCIM solutions.
For example, DC operators faced with an uncertain economy and ever-increasing demand for consistent and improved network services are struggling to maximize both network uptime and profitability, says Frost & Sullivan analyst Amit Singh.
Here, greater automation and holistic insight into how physical infrastructure is managed and optimized as a strategic business asset addresses one of the toughest challenges in the DC – managing network connectivity.
Such insight is made possible by integrating DCIM and Converged Physical Infrastructure Management (CPIM) capabilities to deliver real-time visibility into the physical cabling infrastructure that connects IT assets.
“By enabling deep visibility into all aspects of a DC’s operations, DCIM technology is opening tremendous opportunities for DCs to become more manageable and efficient,” says Eric Woods, research director with Navigant Research.
This visibility paves the way for CPIM’s capacity planning and analytics capabilities to offer a single source of truth for physical infrastructure across assets, power, space, connectivity, network services and HVAC.
An area of significant progress is intelligent connectivity management. DC operators combine CPIM with real-time asset and connectivity information fed by infrastructure management tools to identify and eliminate unused network and patch panel ports and other patching inefficiencies. They can quickly pinpoint and resolve connectivity issues, such as unscheduled changes, that can cause downtime.
Capacity planning and predictive analytics
Advanced analytics are clearly important to the long-term success of DCIM implementations.
The 2013 Uptime Institute Annual Data Center Industry Survey reported “very high” DCIM adoption among survey respondents, with 71% of them listing “improving capacity planning” as a top driver. Capacity planning mistakes, after all, are too expensive and too critical to business to be a hit-or-miss game.
To enable careful planning and control and to ensure every member of the IT and facilities teams understands the processes and tasks at hand, leading-edge CPIM solutions assist operators to commission DC assets with interactive 3D visualization. The commissioning engine ensures that work actions are created quickly and then presented to technicians in clear, visual instructions that eliminate confusion or ambiguity.
Based on the customer’s requirements, the CPIM software analyzes the options and produces the results in easy-to-understand 3D visuals, including the right space for the physical asset; power connectivity; and network connectivity.
More importantly, managers can easily see the impact of a move, add, change, removal or failure of a single device in the ecosystem on the performance of other devices and conditions in the DC. They can manage, and optimize these dynamics continuously in real-time 3D visuals.
They can analyze server power consumption against workloads to identify idle, non-peak or peak periods. With this information, they can control power consumption such as allocating more power to critical revenue-generating servers in peak periods or divert surplus power capacity to support workload expansion within the DC infrastructure.
Integration with other systems
Indeed, the next evolutionary step for DCIM is intelligent capacity planning. This requires automating physical infrastructure management functions and integrating DCIM platforms with other systems and solutions on the DC floor.
Data centers require an alliance of multiple vendors to help manage the entire infrastructure so it is important to facilitate the exchange of information between DCIM, CPIM, IT service management, facilities, building management systems and energy management systems.
For better DCIM integration, the point-and-click asset reconciliation tool of a solution like the iTRACS CPIM allows users to quickly clean up discrepancies and eliminate system-to-system conflicts when combining data from multiple enterprise systems.
At the same time, an improved flat file integration agent offers bi-directional data exchange for plug-and-play integration between CPIM and third-party systems dealing with asset or change management and operational data feeds.
The data feeds can be analyzed to aid faster provisioning of DC assets; higher staff productivity; and better energy management.
Holistic insights from predictive analytics as well as capacity planning capabilities enable DC professionals to maximize capacity, control energy costs and utilize IT assets more effectively. But they will need DCIM offerings that are part of a broader solution that includes capacity planning and predictive analysis, as well as power and cooling management.