Special report on the 9th China Data Center Confer

Special report on the 9th China Data Center Conference 2016

on April 15, the annual China Data Center Conference arrived as scheduled. As the host, this is the ninth consecutive year that China Computer News has hosted the China Data Center Conference. With the theme of smart data center, this conference comprehensively expounds the construction and operation of smart data center from the perspectives of data center planning, design, deployment and implementation, transportation and peacekeeping management, and future evolution, and reveals the way of smart development of data center

with the explosion of information and the rise of cloud computing, big data, IOT, etc., domestic investment in data centers continues to increase. In China, data centers are still just needed. The data center is developing in the direction of large-scale, high-density, intensive, green, intelligent and automatic. Under this trend, the complexity of data center construction and operation and maintenance is gradually increasing, and the traditional data center construction and operation and maintenance mode is facing unprecedented challenges. The construction and operation and maintenance of the data center must get rid of the original rules and regulations, realize the comprehensive transformation from rigidity to flexibility, from extensive to fine, from mainly relying on labor to automation, and build an intelligent data center with high flexibility, high efficiency, high reliability and high intelligence as the basic characteristics

data center operation and maintenance is a new business opportunity

some time ago, I believe everyone was paying attention to the man-machine war between alpha dog and South Korea's Jiuduan chess player Li Shishi. Nie Weiping, the sage of Chinese go, also said in a CCTV TV program that he was very shocked by the man-machine war. Does the data center need artificial intelligence? Maybe your data center can also have alpha dogs

to realize the intellectualization of the data center, improve the availability and energy efficiency of the data center by means of software, and finally maximize the return on investment, which is a problem that every data center user should seriously consider. Zheng Hao, director of solution products of Schneider Electric it division, said

alpha dogs have two magic weapons to win: first, they have the ability of deep learning; Second, it has two brains, one is the drop selector, the other is the chess game evaluator. We can also apply the intelligence of alpha dog to the existing data center. In this regard, Schneider Electric has done a lot of research and practice. Zheng Hao said that this is mainly reflected in the following three aspects. First, in-depth learning, that is, based on a large number of data center construction practice and operation and maintenance management experience, the R & D and manufacturing capacity of end-to-end solutions, and an in-depth understanding of the physical infrastructure of the data center, to form a complete set of basic knowledge base, which is a bit like alpha dog reading chess scores. By using Schneider Electric's real-time data acquisition and tracking system, users can continuously update relevant knowledge and improve their understanding of the data center. Coupled with some analysis tools of Schneider Electric, users can make predictions in advance for the operation and management of the entire data center, or provide decision-making basis for operators. These are the abilities of deep learning

second, decision-making. A large number of equipment in the data center belong to different systems, which are managed and operated by different management teams. Schneider Electric can find the correlation between different equipment through simulation, risk assessment and other methods, and find the problem in time, so as to further reduce the risk of operation and maintenance

third, self optimization. After locating the risk, the key is how to find the best way to solve the problem in metal and non-metal mechanical property experiments like alpha dog's drop selector. This is what self optimization should do. For example, Schneider's intelligent system can detect the downtime of a server in time, and then deploy a virtual machine to take over the business. These are all automated

Amazon AWS (Amazon Web Service) launched its first commercial cloud computing service in 2006. In the process of building and operating cloud computing services, Amazon AWS has accumulated a lot of experience in data center operation and maintenance, and has launched a series of cloud services, realizing the infrastructure, namely, the architecture deployment of code, using code or containers to automatically deploy and upgrade applications, and through the development and operation and maintenance integration (Devops), making continuous development, integration and deployment possible

Amazon AWS is a benchmark to promote it operation and maintenance innovation with cloud services. People are no longer unfamiliar with virtualization technology. Server virtualization or storage virtualization only virtualizes the hardware. On this basis, container technology virtualizes the operating system, middleware, etc. into containers, and then directly puts applications into containers, so that containers can be as convenient as containers for delivery and operation and maintenance. Container technology, Devops and microservices will change the traditional application delivery mode and realize the integration of development, testing and operation and maintenance

Amazon has thousands of teams, and each team is like a startup company. These teams adopt the micro service architecture to achieve rapid and continuous application delivery, and their applications can run in a variety of environments. Dr. Zhang Xia, chief cloud computing enterprise consultant of Amazon AWS, said that Amazon completes an average of 50million deliveries per year

speaking of innovation in operation and maintenance, Amazon AWS advocates the new concept that infrastructure is code. Amazon AWS' data center operation and maintenance has been completely automated. Relying on machines and software to realize the deployment of the whole system has many obvious advantages over manual operation: machines have better scalability, replicability, auditability, and can ensure the consistency of the environment, as well as the security of records. From code development to testing, configuration, deployment and continuous monitoring, Amazon AWS has launched a series of tools for this process, and summarized a set of standard practices, easily realizing continuous integration and continuous delivery

in the process of IT industry development, it service is a new turning point, and cloud computing itself is also an IT service

there are different modes of cloud, which can be private cloud, public cloud or hybrid cloud. What has changed from traditional IT operations to cloud operations? In traditional IT systems, an operating system is installed on a server, on which different applications are deployed. Traditional IT operation and maintenance only need to be aimed at a single server. In the cloud data center, a cloud operating system is installed on hundreds of servers, which is in charge of all it hardware equipment and even the physical infrastructure of the data center. There may be multiple virtual machines deployed on a physical server. At this time, if there is a problem with the physical server, it is impossible to take this server offline as simple as traditional IT operation and maintenance, but we should be very careful, because it leads to the overall situation. For example, when there is a problem with the power supply of a server, the administrator may have to perform multiple checks and operations before replacing the broken power supply. Tang Tao, general manager of technology operation and maintenance of century Internet blue cloud division, said that cloud operation and maintenance is a huge challenge for cloud applications. If we compare traditional IT operation and maintenance to managing a small hotel, then cloud operation and maintenance is equivalent to managing a five-star hotel, and it is chain. Yunyun maintenance has a set of standard execution processes. Users should be cautious every time they deal with it, because it involves cloud applications, cloud operating systems, it equipment and other issues

at present, there are more than 400 engineers in the blue cloud division of 21vianet in China, which can provide international professional operation and maintenance services, including cloud, infrastructure, technical support, customer support, business support and other services. We can provide a cloud platform for localized operation and maintenance according to the needs of local users in China, while ensuring the openness and flexibility of the platform. We can save six copies of backup data in three cabinets in Beijing and Shanghai data centers to improve the reliability and availability of data. Even if five cabinets are powered off, the availability of data will not be affected. Tang Tao said

effective governance of data center

storage used to be one of the bottlenecks of data center infrastructure. Nowadays, the storage market is changing, and new concepts such as software definition, full flash memory, distributed storage and big data storage emerge in endlessly. What kind of storage solution is the most suitable for cloud data centers

in order to understand the changes in the storage needs of data center users in the era of cloud computing, tongyoufeiji conducted a survey and found that data-driven orientation has become the core needs of users. The Information Department of enterprises is changing from a lubricant providing technical services to a propeller guided by data drive and promoting business innovation

based on this understanding, Beijing tongyoufeiji Technology Co., Ltd. (hereinafter referred to as tongyoufeiji) adheres to the idea of taking data as the core. In the storage architecture design of cloud data center, according to the characteristics of data business, it is divided into business service platform, business operation platform, infrastructure platform, storage infrastructure, big data storage equipment Data protection equipment, etc. constitute the storage system of the cloud data center. Such a composite storage scheme can not only ensure the normal access of data, but also ensure the security and integrity of data in case of system exceptions. Tongyoufeiji can help customers build hybrid big data storage platforms in the process of building cloud data centers, and provide dynamically adjustable data storage resources for customers' operation services, business systems and management decision-making platforms through intelligent perception engines

Zhou Lingyun, product marketing director of tongyoufeiji, said that tongyoufeiji's cloud data center storage solution is integrated and unified, which can ensure the connection of storage and realize the transparent migration of data

in recent years, China has frequently issued relevant requirements and specifications on data center energy conservation. This is mainly because the overall energy consumption of the data center is very high, and the power consumption of a large data center is even greater than that of a park as a whole. The overall architecture of the data center is usually composed of IT system, refrigeration system and power supply and distribution system. Because refrigeration consumes too much energy, the overall efficiency of the data center is low

in March 2015, in order to implement the requirements of the "opinions of the State Council on accelerating the development of energy conservation and environmental protection industries" and comprehensively improve the level of energy conservation and environmental protection in data centers, the Ministry of industry and information technology, the State Administration of agency affairs and the National Energy Administration decided to carry out the pilot work of green data centers, aiming to establish a model of green data centers and establish an evaluation system for data centers. Green is the development direction of data center

Lu Tianwen, Secretary General of the China data center energy conservation technical committee, said: in addition to the operation and maintenance of the data center is a new business opportunity, energy conservation is also a huge business opportunity. Our statistics show that in 2015, the scale of China's data center energy-saving market reached 3billion yuan, mainly including the transformation of cabinet microenvironment and the use of fresh air refrigeration system. According to the current development trend and the support of national policies, the scale of China's data center energy-saving market will reach 8.5 billion yuan in the next three years, with a very fast growth rate

in 2015, there were 66 major data center downtime events in the United States, of which 25% were caused by the failure of traditional UPS system. In addition, information security is also a major reason for data center downtime. in order to