| Summary: | A cache appliance is a network terminal which provides cache memory functions, such as object queries service from a user; such objects could be stored in one cache or in a cache hierarchy, trying to avoid carry out service from an origin server. This cache appliance structure improves network performance and quality of service. These appliances use ICP protocol (Internet Cache Protocol) to support interoperation between existing cache hierarchies and web servers, through implementation of a message format to communicate web caches. One cache sends an ICP query to its neighbors. The neighbors send back ICP replies indicating "HIT" or "MISS". When one cache faces an excessive traffic situation, that is, a very high number of service queries from users, ICP protocol may allocate the service to cache which has the desired object. Because of traffic conditions, specific appliance may congests and the requests may be refused, which can decrease network's quality of service. So, a system designed for optimizing cache allocation, considering factors as traffic and priority could be useful. This paper presents a fuzzy inference system design, which uses entries such as number of queries over a time interval and traffic tendency, and as output, the web cache allocation decision that will provides the service; this design is proposed to optimize allocation of caches into a hierarchy for network services performance, so balancing out requesting among hierarchy members and improving services performance.
|
|---|
