Three Hurdles to Data Center Reliability and Efficiency

The pressure is on data center operators to always be “on,” as in avoiding costly downtime and optimizing resource usage.

Downtime can cost data centers hundreds of thousands — sometimes millions — of dollars, and the stakes are only getting higher as the digital transformation forges on. Many outages (43% according to the Uptime Institute) are caused by power issues.

With the added pressures of greater electricity needs, higher power costs and the global movement to reduce greenhouse gas emissions, it’s more important than ever that data centers have the ability to monitor and manage power quality along the entire power chain. A few common hurdles will hamper efforts to operate with the highest reliability and efficiency. Let’s take a look at three of the most common hurdles.

1. Inability to see granular power quality data increases downtime risks. Power quality is generally assessed once when a data center is being brought online or after it’s had a major overhaul. As a result, there is no understanding of how power quality issues are affecting equipment on a daily basis, or how standard operations may be causing power quality issues. Because they operate on non-linear loads, the type of equipment in data centers is susceptible to power quality issues. Those types of loads can lead to harmonic distortion, a common power quality issue that can cause equipment malfunction or failure. Being able to see those kinds of issues at the device level empowers facility managers to take steps to ensure they are reduced and their impact mitigated or avoided entirely.

Raritan_Harmonic_Distortion_Tab_RAM_HR_0

Image 1: Total Harmonic Distortion

2. It is difficult to optimize for efficiency and scale infrastructure without power quality metrics. High-density IT infrastructure cannot be taken advantage of because of stranded capacity and other power quality issues, making it so data centers cannot get the most out of their space and usage. High-level monitoring also can reveal the locations of “zombie servers,” which eat up electricity when they’re not even in use. Being able to see granular data at the rack and device level in real time helps identify and resolve these and other problems that could otherwise lead to equipment failure or other causes of downtime.

Raritan PX4 Web GUI Image_dashboard_0

Image 2: Advanced Power Monitoring and Metrics

3. A challenge with a lot of intelligent rack PDUs on the market today is that they don’t integrate seamlessly with building management systems (BMS) or data center infrastructure management (DCIM) solutions. This means additional tools and features must be implemented to achieve greater efficiency and uptime. The disconnect between the power monitoring equipment and the tools used to manage facilities makes it harder to manage usage as effectively as possible. PDUs should be able to communicate with the BMS and DCIM already in place via universal APIs.

Raritan_PX4_PDUE7.colors6-transparency_0

Image 3: Raritan PX4 Intelligent Rack PDU

An innovative intelligent rack power distribution, monitoring and control solution will be the best way to overcome these hurdles and boost uptime at data centers. It will deliver real-time information, empowering facility managers to make the best decisions for efficiency and optimal operations. The costs and other hassles of downtime — not to mention increased power demand, higher costs, and environmental concerns — are too large to continue operating with the status quo.

Raritan, a brand of Legrand, has launched the forward-thinking PX4 Intelligent Rack PDU, which provides data centers with unrivaled visibility, reporting, security, and alerting, helping to drive operational excellence. Discover more about the PX4 PDU.