摘要： 实时对网络海量信息截取、分析和处理的负载远远超出了单台计算机的处理能力,文章采取在Linux操作系统内核中将IP报文截取,并按照转发策略对IP报文进行重新定向发送给下游处理机的方法,在操作系统底层实现IP报文的分流,满足了对网络海量信息进行实时分析和处理的技术要求。

关键词: Linux网络协议栈, IP报文, 分流

Abstract: It goes beyond the capability of the single computer to capture, analyze and treat with large information transfered on Internet at real-time. The method that IP datagram is captured and diverted in the kernel of Linux operating system to subsequent processing computer is presented in the paper, which meets the technical request to analyze and react those information at real- time.

中图分类号: 

• TN915.04