Since the 1960s, data centers have become the most important part of an enterprise’s IT infrastructure. In the 1990s, with the development of cloud computing, enterprises began to try to handle computing tasks through distributed computing and realize the instant sharing of computing storage. In recent years, Internet of Things users continue to grow, artificial intelligence and other technologies continue to develop, distributed edge computing has become a key support for enterprises to win the digital era.

Extending to cloud providers, colocation facilities, edge devices, and edge computing, the boundaries of the generalized data center are blurring, with increasing complexity – performance, reliability, compliance, security, and more – making hybrid computing environments more difficult to manage. Among them, how to deal with the operation and maintenance management of edge computing has become an important challenge for enterprises.

Recently, Guo Feng, Deputy Secretary General and Senior engineer of the Energy Conservation and Emission Reduction Promotion Committee of the Chinese Society of Electronics, Yang Weikai, Chief Operating Officer of the Technology and Operations Department of DBS Bank (China) Limited, and Sun Boya, digital product Manager of the key Power Business Department of Schneider Electric, visited the “Future Talk” section of the data center. The topic of “Efficient operation and maintenance management of edge computing data center” is analyzed and discussed.

Collaborative evolution from cloud to edge

Now the cloud to the edge has gradually become a new trend of enterprise deployment, “octopus” is a very image metaphor of the cloud to the edge model in the industry, the octopus body is distributed with neurons, the brain accounts for 40%, the remaining 60% dispersed on the legs, this 1 brain total control coordination +N cerebellum decentralized execution structure, and the center cloud + edge architecture is very similar.

In this architecture, the cloud and the edge each play their roles and interact. With powerful data centers, cloud computing provides elastic computing, storage, and network capabilities for service applications, and is applicable to non-real-time and long-term service scenarios. Edge computing focuses on real-time, short-cycle, local decision-making and other business scenarios, such as audio and video live broadcasting, the Internet of Things, meta-universe, etc., sinking the workload to the place away from the terminal device or close to the end user, achieving lower network latency and improving the user experience.

5G and Internet of Things technologies have driven a series of demands for high data processing delay requirements such as visual recognition and automatic driving, and traditional data centers cannot respond in time to handle these services. Guo Feng believes that edge data centers can well help data processing services with high latency requirements.

As a global leader in data center infrastructure and digital services, Schneider Electric classifies edge computing into three categories based on infrastructure environment characteristics: The first category, IT environment, that is, specialized IT or computing equipment space, such as server rooms, power distribution rooms of enterprise networks, small data centers, etc., these Spaces are basically built in accordance with standard computer rooms, with good environment and cleanliness; Second, the commercial office environment, such as office areas, warehousing and logistics centers, may have problems with temperature and humidity, noise, dust and other aspects; The third category, industrial and harsh environments, such as production lines, instrumentation rooms, etc., most of this scene equipment is exposed to the factory, and the environmental physical threat pressure is larger.

“These are typical edge computing scenarios that pose operational challenges due to their dispersed distribution, unfriendly environments, and unattended presence.” Sun Boya further added.

As a representative enterprise of edge computing deployment, DBS Bank is the first foreign bank in China to fully adopt iot sensors in its IT room with independent operation and maintenance. DBS Bank is a leading financial services group in Asia with a presence in 19 markets. It established its China office in Beijing in 1993 and became one of the first foreign banks to incorporate in China in May 2007. DBS China branch network and business has grown rapidly, and there are now 12 branches and 22 sub-branches in China.

Adhering to the principle of “data speaks for all management decisions”, DBS Bank has carried out multiple exploration practices in digital upgrading. In recent years, DBS has continued to explore the use of innovative technologies and digital drivers to improve productivity, and focus on reducing energy consumption and protecting the environment in daily operations. By 2023, DBS has digitized the management of all its computer rooms by installing sensors, and benefited from risk management, environmental sustainability and procurement optimization through edge computing.

Technology to address edge computing operation and maintenance challenges

While most businesses already see the benefits of edge computing, resilience, security, and sustainability remain important challenges for hybrid environments.

“Edge computing is generally close to the business, first of all, the environment is a relatively big challenge, second, the increase in the scale of operation and maintenance, the professional requirements of operators are more stringent, and finally, the load of edge computing changes greatly, which also puts higher requirements on the control capacity of infrastructure.” “Said Mr Guo.

At the operation and maintenance level, the most common problems focus on increasing infrastructure complexity, 100% availability requirements, limited resources, multiple monitoring tools, and multi-site distribution

In this regard, DBS Bank encountered the most immediate operational challenges, but also from the manpower, risk control, and sustainability. Yang Weikai said that the distribution of the network by human inspection is not only inefficient, once the risk occurs, the response will not be timely, through the edge computing nodes can be better and more scientific management.

In view of the typical problems of manpower, professional knowledge, and many suppliers, Sun Boya believes that various roles, including end users, equipment suppliers, operation and maintenance service providers, need to use digital management tools, and the formation of ecological collaboration is the key to crack the problem.

Previously, through cooperation with Schneider Electric, DBS Bank completed the deployment of 477 iot sensors in branches and sub-branches in early 2023, achieving three improvements:

First, reduce risk and increase productivity. For banks, security and risk management are critical, and with Schneider Electric EcoStruxure IT Infrastructure Specialists (ITE), DBS has effectively avoided the large number of on-site inspections caused by false alarms. In the past, traditional environmental monitoring systems (EMS) could only tell if a threshold alarm had been triggered. Now with iot sensors, employees can monitor these metrics and their history and trends in real time to determine if these are indeed true alerts. In the first three months of operation with the new system, DBS successfully identified and filtered out more than 70 per cent of false alarms.

Second, improve sustainability. In the past, in order to ensure that temperature conditions do not exceed alarm thresholds, some field users may overcompensate at temperature setpoints, setting the temperature well below the actual required temperature, resulting in the server room being kept unnecessarily cold, and users can only correct these conditions when they visit the site during regular preventive maintenance programs. Thanks to Schneider Electric’s IT Infrastructure Experts (ITE), users can monitor on-site temperatures in real time and correct any overcompensation in a timely manner to avoid energy waste.

Third, procurement and cost optimization. Sensors also play an important role in extending the service life of equipment, for example, by collecting data on the actual load of the UPS at each site, DBS is able to determine the appropriate UPS power based on the collected data for operational purposes without causing overcompensation. At the same time, this also helps banks better plan the real estate requirements of the scale of IT server rooms.