On Wed, 2009-05-27 at 14:11 -0600, John Kennedy wrote: > The archives also seem to indicate that a certain amount of port > specific tuning for TCP is necessary. So it would be helpful if > someone could suggest some standard LWIP test scenarios along with > some guidelines of what to tweak to determine the best configuration > for one's particular needs. Otherwise it's a lot of trial and error.
The problem here, for us, is that there is no "standard" for test scenarios or benchmarks for TCP. What might improve things for one workload could make performance worse on another, and this is why it's left to the person who knows how lwIP will be used to determine what is best for these tunable parameters. I can appreciate that this requires that the person who knows how lwIP is being used (i.e. you!) also knows a bit about TCP and how the parameters affect network performance. Documentation is probably the best we can do to improve matters, and improvements are very willing accepted! However, I don't really want to see lwIP become a reference for what generic stuff like (for example) an MSS is and how its size affects things; the documentation should be lwIP specific. Kieran _______________________________________________ lwip-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/lwip-users
