Enterprises are often challenged by the complexity and fixed nature of their aging data centers. We commonly see the lack of performance capacity necessary to handle the goals of advanced “Industry 4.0” application scenarios and reduced energy consumption. To address these challenges, organizations are increasingly moving to Smart Data Center solutions that deliver the agility, reliability, and efficiency that support the rapid service deployment needed to compete and win in this new era.

Challenges of Legacy Data Centers

Legacy data centers typically lack the operational reliability achieved by the current generation of equipment, and as a result have generally depended on redundant systems for stability. In addition, the complexity of managing data center growth has increased in response to new business demands. Issues include:

• Non-standard designs make many legacy data center operations very labor intensive to maintain, and unsustainable for meeting the fast-rates of change in today’s business climate.

Example: Tencent, China’s largest Internet service portal, required over 400 days to gain the first 100 million users for its WeChat instant messaging application. It only took an additional 86 days to achieve its 500 millionth user.

• The excessive energy use of older data center technologies cuts into corporate profits and adds to each operator’s carbon footprint.

Example: The 2013 Uptime Institute annual data center industry survey documented thatthe annual average Power Usage Effectiveness (PUE) of legacy data centers is about 1.8 — and on the order of 50 percent of the energy used creates no discernible business value. Globally, annual data center power consumption has reached 300 billion kWh — approximately 2 percent of the world’s total electrical use — and is the source of emission for 299 million tons of carbon.

• Data center outages have serious consequences in terms of lost revenue, lost productivity and damaged reputations.

Example: Alibaba Group earned USD 9.3 billion in sales revenue during last year’s annual Double-Eleven Crazy Shopping Festival. The transaction rate through Alibaba’s eCommerce platform averaged USD 6.5 million per minute.

Architectural Simplicity and Ease of Use

Building, data center infrastructure, and IT management systems are all operated independently in traditional data centers. Huawei’s Smart Data Center platforms are based on a unified management console able to address independently functional modules for meeting a wide variety of business requirements with flexible configurations. Built to be simple and scalable, our current generation of data centers accommodate rapid changes in customer demand. Building complete data centers now takes just weeks, instead of the months and years that had been required by traditional data centers.

Optimizing Efficiency in Future Data Centers

Huawei’s Smart Data Center energy-management modules are driven by sensor data captured at key nodes. The result is an optimized energy consumption profile for all systems and devices:

• iPower: Power supply policies manage all devices to ensure the highest efficiency and load balancing.
• iCooling: A centralized system manages air conditioner groups through an intelligent controller. Instructions are sent based on field conditions — temperature and humidity data — to ensure optimum cooling efficiency.

Example: With these new features, the annual average PUE of an Smart Data Center based in Hannover would be reduced to 1.42 — prior to accounting for free cooling. This means that a legacy data center with a PUE of 1.8 would save USD 1.66 million per year (at USD 10 cents per kWh) for a 10 MW-capacity data center at 50 percent load.

According to the Uptime Institute’s Abnormal Incident Reporting (AIR) database, human error accounts for 70 percent of unplanned data center downtime. These failures can be avoided by using the automated policy management resources included in a modern data center solution. Smart Data Center minimize the risk of operational failure with:

• Standardized construction and maintenance
• The use of proactive, self-optimizing management protocols to detect and rectify risks prior to failure

The use of distributed sensors throughout modern Smart Data Centers provide an effective new means for collecting and analyzing raw data for use by Operations and Maintenance (O&M) staff. The sensors not only provide accurate, reliable data, but also free technical personnel to focus on supporting core business activities rather than chasing old problems.

Smart Data Centers for the Industry 4.0 Era

Data centers of the future will incorporate large sensor networks to monitor the environment and ensure operational continuity and energy efficiency, while Big Data analytics for performance monitoring will enable fast and accurate forecasting of data center issues. The goal in all cases is to enable IT staff priorities to focus on staying ahead of persistent problems and to better plan for long-term sustained operations.

Through these intelligent, automatic and self-managed data centers, organizations have new powers to create and deliver IT services with the speed, reliability, and efficiency demanded by their customers.