Conversation maintaining method and device

Abstract

The invention provides a conversation maintaining method and device. According to the conversation maintaining method and device, a conversation maintaining identification is set for a client side through load balancing equipment, the conversation maintaining identification is sent to the client side through an HTTP redirected response message which contains the conversation maintaining identification, the client side carries a conversation maintaining message in a subsequent HTTP request message, the load balancing equipment sends the HTTP request message to a corresponding server according to the correspondence between the conversation maintaining identification which is locally stored and servers so that HTTP request messages containing the same conversation maintaining identification can be sent to the same server, then conversation maintaining is achieved, and the server sends an HTTP response message to the load balancing equipment after receiving the HTTP request message. According to the conversation maintaining method and device, due to the fact that the HTTP response message can be directly transmitted to the client side by the load balancing equipment, the correspondence between the conversation maintaining identification and the server can be obtained without the need of analyzing each HTTP response message sent by the server. Accordingly, burden of the load balancing equipment is relieved, and processing capacity is improved.

Description

technical field [0001] The present invention relates to computer technology, in particular to a session maintaining method and device. Background technique [0002] In the application scenario of a server cluster, multiple servers with the same or similar functions are connected together through a load balancing device. When a client requests a service from a server, the load balancing device transfers the client's Hypertext Transfer Protocol (Hypertext transfer protocol, hereinafter referred to as: HTTP) request message is sent to a server in the server cluster, that is, according to the current load of each server in the server cluster, the HTTP request message of the client is reasonably allocated to different servers, In this way, each server can share the load in a balanced manner and improve system performance. In this case, different HTTP request messages of the same client may be assigned to different servers for processing. [0003] However, in many applications, ...

AI Technical Summary

Problems solved by technology

[0004] However, in the prior art method, the cookie information is set by the server, and the load balancing device cannot predict which HTTP response message the server will include the cookie information in. Therefore, it is necessary to parse each HTTP response message sent by the server to obtain the cookie information. The corresponding relationship between information and servers reduces the processing capacity of load balancing equipment

Images