本文分析ipvs的各种模型的区别与优缺点,并基于虚拟机实现简单的负载均衡集群,并且完成简单负载均衡集群的配置和其配置的方法、流程。
关键词: Web服务器;负载均衡;集群;
Abstract
The rapid development of today's Internet and everywhere in the rapid development of social networks, e-commerce, information portal application has profoundly deepened people's lives and gradually change people's traditional way of life, and update people new awareness of the network, the greater extent do stay at home can handle a lot of transaction. However, with the concurrent growth in the same enterprise-class server to bring more and greater challenges. Is now a Web server to be the face of the flow can not be compared before access to more users of the service, to provide a more stable service, faster response and processing capacity.
Web server's performance depends on the server hardware and software for the new challenges, including CPU, RAM, system OS. When a lot of traffic to access the server due to performance constraints, delayed response, packet loss, or result in data delay retransmission serious occupied bandwidth request further increase the burden on the server, the possibility of serious downtime caused by loss of all responses. But with the increasing demand and growth, a single upgrade server equipment is far from a permanent solution, the high cost of inputs as well as the upgrading of existing waste of resources need to be taken into account. Even the server update to the latest and best equipment is not enough to meet the business requirements. The most common solution for this challenge is the server cluster load balancing.
Load balancing is the most scientific technology to solve this challenge. Web server load balancing technology system consists of multiple Web servers at the same level in a symmetrical manner to form a collection of servers, each server has the same status, and can provide service does not require the aid of other servers, through the Director to the uniform scheduling structure on a single server. Receiver to the requested separate server to respond to customer requests. Such a system processing load assigned to a different server to provide a solution to a large number of concurrent users to access the service parallel processing is load balancing.