On 10/31/11 11:11, Stuart Henderson wrote: > On 2011/10/30 15:41, Daniel Melameth wrote: >> >> sthen@ created a patch a while back that addresses >> this--http://ns2.spacehopper.org/openbsd/base/altq_tbradapt.diff--but >> I have not used it in a while. >>
Thanks for this input. That's exactly what I wanted to do. > > That diff needs rewriting to be configurable from userland rather than > just having a few preset options in the kernel, I'll love to do this. > but seeing as altq is > being replaced I think it makes sense to wait until the new prioritisation > code can limit output to a rate less than the interface bandwidth. Hmm, but when will this happen? Henning wrote something about a 5-year plan when he introduced the prio queueing ?!? Should I really wait for that? My thought is to add an option "overhead" to altq that specifies the size of additional headers of lower level protocols. A sane default for plain ethernet would be 7+1+12=20bytes (Preamble, SOFD, Interframe gap). (Correct me if I'm wrong here). Then a "minsize" option is needed for ethernet where ethernet frames are padded to be no smaller than 64 byte. For PPPoE an additional unit "cells" for the "bandwidth" option (or a flag "cellular"...) could be used to specify that the packet size (including headers) should be rounded up to a multiple of the ATM payload size (48 bytes) and an according amount of ATM header bytes be added. Is this a sane design?
