| With the development of the Internet in recent years,the explosive visits,hundreds of millions of requests,and huge system integration has brought huge challenge to establishing a high-quality network service system.Through intelligent algorithm and various excellent middleware,a high-quality web site can quickly respond to the concentration,analysis,distribution and processing of request,is equipped with disaster recovery capacity for all kinds of malfunctions and disasters,so as to provide users with good service experience.The emerging distributed idea and related technologie in recent years provide a reliable platform for building high-performance web sites.In this paper,the author aims to design and realize a high-performance web site that vertically and horizontally separate system layers of different functions to realize data distribution,liberate the performance bottleneck of single points and improve the system's concurrent and throughput capacity through technical means of flow control,load balancing,distributed services,and distributed database etc.The system architecture is mainly composed of four parts: application layer,business layer,data layer,and cache layer.The application layer mainly deals with the operation and maintenance tasks such as flow control and load balancing.It also provides hot standby to ensure the IO processing capability in busy hours.For the business layer,which designed by a service-oriented architecture,it horizontally expands business processing capabilities and provides services scheduling,service load balancing,service monitoring and other functions by adding a certain number of service providers.The underlying data layer is responsible for reading and writing data across the entire web site.The addition of separated reading/writing functions can solve problems such as data congestion and IO bottlenecks.Distributed databases can improve data processing capability through horizontal sub-division and vertical slicing techniques.The cache layer efficiently increases data reading speed and reduces redundant database operations by building page-,service-,and data-layer caches.Finally,it is discussed in this paper the high concurrency control issues,extension and disaster recovery performances of this system.It is also organized relevant tests to monitor the performance loss,and determine whether the performance index meets the system requirements.Following a distributed idea,this system aims at digging deeper into the performance requirements of a high-quality service site and providing and implementing a set of distributed system design schemes in which each layer can perform performance cutting and scheduling.Sufficient tests are also made on system reliability and availability to launch a highly adaptable,stable,reliable and easy-to-use web site. |
