On 10/17/07, Girish Venkatachalam <[EMAIL PROTECTED]> wrote: > Dear friends, > > I hope you benefit from these articles. > > I am discussing how I improved the performance 12 x (1200 %) of an > existing file transfer mechanism. > > High performance I > http://www.ibm.com/developerworks/aix/library/au-highperform1/ > High performance II > http://www.ibm.com/developerworks/aix/library/au-highperform2/
Hi Girish, There are a couple of inaccuracies/omissions in the articles. 1. The first part of the article mentions <quote> However, there are influences that occur deep down at a network level. You've probably heard of IP fragmentation. Mainly, it hurts performance. Fragmentation and reassembly are a costly affair and, though this is performed only by intermediate routers, it has severe impact on throughput. </quote> This is wrong as in IPv4 fragmentation can be performed by a router between the source and destination of an IP datagram, but reassembly is only done by the destination device. you can refer Comer's book for this. Another online source is http://www.tcpipguide.com/free/t_IPMessageReassemblyProcess.htm 2. I am not sure if you plan to cover this in part 3 but there is no mention of kqueue or epoll in either of the articles. These can improve the performance of networking programs significantly in my experience. -- Vinayak http://www.linkedin.com/in/vinayakh _______________________________________________ bsd-india mailing list [email protected] http://www.bsd-india.org/mailman/listinfo/bsd-india
