面向大型网络的多管理对象信息的组预取算法

如何减少网络管理系统对网络带宽的占用是网络管理研究的难点之一.结合管理信息库(management information base,简称MIB)信息的收集策略直接影响着网管系统性能的特点,针对异构大型网络环境下多管理对象的信息收集问题,提出了一种从多个被管设备中提取有关MIB信息的新算法--组预取算法(group-prefetching algorithm,简称GPA).该算法能在不改变现有的网络管理协议的情况下,将多个分散的、在未来一段时间内的MIB对象的访问组合在一个预取组中进行访问,减少对管理对象的访问频度,从而可以较少地消耗所用的网络带宽和CPU资源,减小整个网络开销.

A Group-Prefetching Algorithm of Information for Multi-Managed Objects in WAN

It is increasingly a difficult problem of how to decrease the usage of network resources by network management systems. Otherwise, the performance of network management is affected by the scheme to collect MIB (management information base) information. A new algorithm, GPA (group-prefetching algorithm), is put forward, which is used for retrieving MIB information from the managed devices in heterogeneous WAN (wide area network). GPA can adjust many little retrieving objects into one prefetching group to access,decrease the frequency of retrieval,and thus can use fewer network resources and decrease the network overhead without alteration of the existing network management protocols.