www.91蝌蚪I在线看h网站I亚洲黄色特级片I欧美做爰猛烈床戏大尺度I五月婷婷六月丁香I国产免费一区二区三区四在线播放I日本精品视频在线播放I日韩免费片I午夜av成人I黄色a级大片I国产尻逼I妺妺窝人体色777777I伊人日韩I在线观看91视频I尤物在线精品I欧美洲69xxxxxI欧美午夜精品理论片a级按摩I日韩污视频I精品中文字幕在线观看I台湾少妇伦欲电影I秋霞影院avI日本特级淫片I久久久视I美女撒尿无遮挡网站I日本在线wwwI久久这里只有精品6

To give impetus to the detail management of the whole network data flow and get more benefit of running the data management, Henan Mobile cooperates with Sugon Company to conduct the management and analysis project for system flow. The purpose of Henan Mobile resource pool project is to build business supporting cloud for internal sharing, which is based on the business computing needs for the future. According to the competition situation and users’ requirements, it will realize real-time billing and reminder of data flow, optimize billing system for data flow, lower the tariff level gap with or without the package, improve the management ability, continuously improve customers’ satisfaction, and greatly reduce the complaints towards flow.

This platform will be used in the data management in the first place. The speed of data processing is based on the platform, proving data support for the development of company’s business. Secondly, the platform will also deal with monthly bills and replace the existing traditional Oracle database group, small-scale machines and high-end FC disc arrangement model, which could temporary ensure the providing of users’ bills on the first day of each month. However, with the data’s increasing speed of 15% each year, the existing system will not be able to satisfy the requirements of rapidly developed business, according to the client’s prediction. That’s the reason why they want to establish a higher performance and expansion ability platform to replace the original system.

According to years of accumulation and researches in distributed computing, the Sugon Company recommends distributed computing platform plan–Hadoop, which is based on IB network, to solve the bottleneck problem of network performance.

This project uses 150 A620r-G dual servers to establish Hadoop platform, and 100 A840r-G four-server servers to build database processing platform. The entire system is connected through Infiniband network. In the design of IB network for the system platform, all connections for network equipment adopt redundant design. This system achieves the Tri-networks integration among Ethernet, Fibre channel and Infiniband, through the usage of the former plan to meet the high-speed communication requirements of Hadoop group and database processing platform.