Big data increasingly highlights the important value

Big data is playing an increasingly important role in the information society, promoting the transformation of the commercial society and creating a new type of consumer business; however, massive, unstructured, and big data that needs fast real-time processing also brings to the traditional data center. Great challenge. Recently, many experts called for reconstructing the data center in the era of big data at the "2013 Information and Communication Industry Data Center and Big Data Summit (3rd), hosted by the People's Posts and Telecommunications Newspapers and the China Information Industry Network, and organized by Epwer. Rebuilding business ecology."

Big value of big data: expanding the new engine of information consumption

Data has become the core asset of modern enterprises and has become the mineral and “oil” of the information society. The age of big data for large-scale production, sharing and application of data is approaching. Qu Ying, Vice President of China Division of Ned Electric's IT Division, introduced that “the quantity unit Z represents the 21st power of 10, and the total amount of water on the Earth is 1.4Zg, and the total amount of air on the entire planet is about 0.6ZL. In 2012, all global data reached 2.7ZB. By 2020, the global data volume will exceed 40ZB, this figure will be 14 times the end of last year. In 2020 China's data volume will exceed 8ZB, its growth rate exceeds The global growth rate is 23 times that of 2012."

Data from the DCCI Internet Data Center shows that 60 Instragram photos are uploaded every 1 second; every 1 minute, 60 hours of video is passed to Youtube; every 1 day, the number of logs generated by search engines is 35T. The amount of information that has been generated so far is 5 exabytes (5 billion GB), and the amount of data generated in the past three years is even larger than the amount of data in the past 40,000 years. According to Chen Luying, chief technology officer of Beijing Broadcom High-tech Development Co., Ltd., “Data has become a core asset of the company. Control over the data determines the right to control the market. Enterprises that are closer to the end user will have greater say in the industry chain. Those companies that have a large amount of effective data and companies with strong data analysis capabilities will play an important role in the big data industry chain. The key is who has more, more accurate, and more valuable data."

Wu Suining, editor-in-chief of the People's Posts and Telecommunications News, pointed out that big data brings great value. In the era of Internet and big data, the business ecosystem has undergone earth-shaking changes. Everything people can imagine about human activities is almost being redefined. In order to deal with market competition and satisfy user needs, operators should attach importance to data resources as much as they regard network resources and user resources, fully tap the commercial value of big data, and become operators of big data.

Director Xie Yuqi of the Policy and Standards Division of the Department of Communication Development of the Ministry of Industry and Information Technology stated that under China's current environment of promoting information consumption and stimulating domestic demand, big data will become a new engine for expanding information consumption, and will be more accurate through the collection and analysis of big data. Insights, making more scientific predictions and decisions that can provide new information and communication services. For example, the use of big data analysis to rationally plan traffic, monitor epidemics, help the government to innovate in social management, link real-time data processing and analysis with real-time advertising services, and help customers by excavating user behavior data. Conduct personalized recommendations and precision marketing. These measures have innovated the information consumption business and business model, and have greatly increased the attractiveness of information consumption.

Big Data challenge to build a green, smart and secure data center

Big data is the most dazzling new business in the evolution of information and communications. In the future, there may be more new technologies and new services. However, the development of any new technology and new services cannot be separated from the high level of information and communication infrastructure. Facilities, especially the data center. The energy consumption of data centers is now a global problem. Surveys show that the growth rate of energy consumption in global data centers in the past five years was 56%. The demand for data center traffic processing capacity in China has grown faster, and energy consumption in data centers has increased. The problem is even more pronounced.

Chi Jiuhong, general manager of Huawei's data center infrastructure product line, said that in the past, due to the lack of a large-scale data center and lack of awareness of energy conservation, the actual ambient temperature in the data center room has been 18 to 25 degrees. It was not until 2008 and 2009 that everyone reached a consensus that the temperature was raised to 27 degrees, which gave tremendous room for the use of natural cooling. For example, Beijing's average annual temperature is 12 degrees. If you increase from 25 degrees to 27 degrees, the time for natural cooling will increase by 17%, and the annual cooling efficiency will increase by 30%. If you use an air-cooled data center, you can use 25 When the degree is raised to 27 degrees, each degree can save energy by 2%. If it is a water-cooled data center, each degree can be increased by 3%, which is a huge energy saving. Google's data center consumes 2.26 billion kWh of energy a year. At present, 50% of its global data centers are completely cooled by nature. The PUE of the top ten data centers is below 1.2.

Director of the Internet Research Center of the Ministry of Industry and Information Technology and Director General of the Cloud Computing Development and Policy Forum, He Baohong, said that the energy consumption of IDC in the entire ICT industry is quite high, accounting for about 1 to 1.5% of the global power consumption. In the U.S. data center, energy consumption accounts for 1.7% to 2.2% of the U.S. energy consumption. Energy consumption in China's data centers is basically between 2.5% and 3%. In the next 5 years, data center traffic in China will have about 7 to 10%. With times of growth, the capacity of the data center will at least double, and the energy saving problem in the data center is very important. To this end, the state has issued a grading standard for green data centers and put forward specific requirements for green management, equipment and building layout. At present, the highest level of green rating for domestic data centers is only 4A, and there is no 5A green data center.

Xie Yuqi pointed out that in the course of the development of data centers around the world in recent years, there have been some blind and overheating issues. To this end, under the conditions of comprehensive consideration of energy supply, natural environment, market demand, land supply, industrial environment, talent support, and transportation, the Ministry of Industry and Information Technology, in conjunction with the requirements of basic network layout and information security, at the beginning of this year, the National Development and Reform Commission and the Energy Bureau, The four ministries and commissions of the Ministry of Land Resources and the Electricity Regulatory Commission jointly issued the “Guiding Opinions on the Layout of Data Center Construction”, and put forward the principle of market demand orientation, the priority principle of resources and environment, the principle of regional coordination and coordination, the principle of multi-factor coordination, and the principle of equal emphasis on development and security. Such as, through the implementation of six months, has gradually formed a consensus in the industry and enterprises, the rational planning of data center construction, the role of achieving resource conservation is increasingly evident. Taking the mega data center as an example, from 2011 to now, there are about 20 ultra-large data centers that are planned to build more than 10,000 standard racks (100,000 servers) nationwide. PUE is basically below 1.5.

Experts also stated at the meeting that the explosive growth of big data has also brought great challenges to the capacity, scalability, and transmission bottlenecks of data center storage systems. Connected to it are server computing capabilities, memory storage capabilities, data center switch transmission capabilities, and so on, all facing new technological challenges. From the perspective of the data center itself, as users continue to grow their demand for big data storage, mining, transmission, and on-demand dynamic allocation of resources, the traditional data center network architecture is facing major changes, and a new generation of cloud computing data centers will It is the direction of upgrade and evolution. Xie Jun, chief architect of the Cisco China Data Center, said that there are now 15 billion terminal devices on the Internet. On average, there will be 2.2 terminals connected to the Internet. This data will reach 25 billion by 2025; The massive network activity has driven us to deliver more information on the Internet, and unstructured data will become the mainstream. Virtualized applications are widely used, and the virtualized nodes have exceeded the nodes of the physical machine. Therefore, Cisco will vigorously promote the borderless network, support the SDN network operation, and establish a common support platform for big data within the data center.

In addition, data center security has also received great attention at the conference. The recent information security incident once again sounded the alarm for us. At present, the security protection capability of our country's public infrastructure is still insufficient. The features and service models of new technologies such as big data and cloud computing have exposed us to various new security risks. . The opening and sharing of IT infrastructures has increased the security protection risks. Mass data cleaning and filtering and identity authentication access have brought about data security hidden troubles, and bad information processing and personal privacy protection also have great challenges. Cheng Wenjie, CheckPoint's China regional technology manager, said that 63% of data centers are currently undergoing consolidation, introducing virtualization technology and switching to cloud computing, which poses a great challenge to the overall security situation. Because the boundaries of the network after cloud computing platform introduction will become very vague, it is difficult to divide the security domain by way of VLan. How to ensure the security of these platforms while being used and dynamically migrated, is very troublesome. problem.

Big Data Ecology: Operators Make Big Data Applications

There are still some difficulties and challenges in the promotion and application of big data technology. Wu said that we need to strengthen planning and design and build a good business ecosystem. On the one hand, grasp the basic data specification, which is the cornerstone of the development of big data. On the other hand, we must make good use of existing data, which is the driving force for the development of big data. China’s population ranks first in the world, and it will become the country with the largest amount of data. However, we do not pay enough attention to data preservation and the utilization of stored data is not high. In addition, some departments and agencies in China have large amounts of data but are unwilling to share it with other departments, resulting in incomplete information or repeated investment. We should actively appeal to break through data separatism and blockade through institutional reforms, pay attention to open information, and pay attention to data mining. At the same time, strengthen the support for big data technology and product R&D, and explore the good business model of big data services.

The operator communication network and operation support system generate a large amount of user behavior information and contact information every day, including the customer's association circle, position trajectory, consumption habits, content preferences, etc. These are the inherent advantages of operators in developing big data applications. With core wealth, operators have also been actively promoting big data applications in recent years. According to Wan Jun, chief engineer of China Telecom’s government-enterprise customer service division, operators’ exploration of big data applications is divided into four aspects. The first is data analysis based on speech recognition technology, and nearly 160,000 call center agents in the domestic outsourcing market. China Telecom has a market share of 60,000. It can analyze the call center's massive voice data, establish a call center assessment system and conduct product association analysis; secondly, based on the intelligent analysis of video data, China Telecom has a million-level camera deployed in the country. In every corner, information can be mined from massive video and public video image analysis services can be provided. Thirdly, network traffic analysis can be carried out. Through the DPI deep packet inspection technology, it can provide targeted behaviors to groups of general internet users. The fourth is the location data analysis, the analysis of the location and trajectory of the mobile communication users, and the probability of effective statistics of the crowd frequency in hotspots. In addition, in the cloud era, China Telecom will deploy a 4+2 structure in the data center. In addition to building cloud computing centers in four original core areas in Beijing, Shanghai, Guangzhou, and Chengdu, it will also build large-scale clouds in Inner Mongolia and Guizhou. Computation center.

Bai Lin, deputy general manager of China Mobile Internet Base, said that big data should solve two problems. The first is the issue of reserves, how to process data reserves well, the second is to solve the problem of how to output, and effectively push data to Real production services go well with well-defined rules and templates. This is a very important prerequisite for the future big data to truly generate business models. The China Mobile Internet Corporation under preparation manages and operates more than 20 businesses in the entire network including China Mobile Fetion, Caiyun, Wireless City, Consonance, No. Book Manager, and 139 E-mail. The monthly active users of MM are nearly 27 million. Fetion and email are almost 60 million times, and Fetion is 80 million times. It will focus on building an open platform to provide capacity sharing from user introduction, operation support, product profitability, and product functions. In June of this year, the issue of Internet access was solved. . China Mobile's core network and Internet services generate more than 10T bills, 30T Internet logs, 100T signaling data, and 100T audio and video and unstructured data every day. For a large number of user network communication behavior data, China Mobile provides developers with products such as “number” and “advance”, integrates the current Internet platform, billing platform, and content of large-scale data to integrate analysis and become the final analysis. Reports are provided to developers.

Kang Nan, Manager of Operation and Maintenance Department of China Unicom Cloud Data Co., Ltd., stated that China Unicom has planned cities in Langfang, Hohhot, Zhengzhou, Dongguan, Nanning, Wuxi, Chongqing, Xi’an and Harbin to start construction of 9 major data nationwide. Central production base. Focusing on the provision of cloud-based capabilities, we must develop features such as nationwide radiation, scale distribution, virtual storage, flexible scheduling, security protection, and green energy conservation. The Nine Data Centers are interconnected through a new generation of high-speed, high-capacity, and high-performance networks to provide secure and reliable protection for urban data and information storage.