Date: Sat, 4 Mar 2000 23:43:07 +0100 (CET) From: Juergen Lock <nox@jelal.kn-bremen.de> To: kuku@gilberto.physik.RWTH-Aachen.DE Cc: freebsd-isdn@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG Subject: ijppp for isdn, ppp compression, and netgraph (also: load balancing) Message-ID: <200003042243.XAA82879@saturn.kn-bremen.de> In-Reply-To: <200003031038.LAA52158@gil.physik.rwth-aachen.de>
next in thread | previous in thread | raw e-mail | index | archive | help
In article <200003031038.LAA52158@gil.physik.rwth-aachen.de> you write: >Don't know whether it is the right term but what would be >desirable is a way to raise priority of certain protocols, and >lower e.g. ftp, http and such. > >You know what I mean? You are fetching a big file, but wanna do >an rlogin at the same time. I'm doing so right now and >it's no fun to type while 7.2Kb/s are running over your head >at the same time :-) I was just looking (finally :) at ijppp again now that it knows i4b, and it already does something like this, look for `set urgent' in its manpage (man ppp). (if you still want to stick with isppp or raw hdlc you could probably also use altq but thats not in the base system.) Of course to really help for the fetching-a-big-file case the same thing would have to be done at the other end of the link as well... And the other reason i'm looking at ijppp is ppp compression. It currently supports deflate (rfc1979) and predictor1 (rfc1978), which should at least help if the other end is running bsd or linux, but if your other end is something like an ascend or an external router (zyxel, cisco(?), there are probably more that speak this protocol), you'd want stac lzs (rfc1974), or if its a wintendo box even you'd want M$' special version of that (yes of course they invented their own `standard' again.) So my question is, is anyone working on this? There is (alpha) code that does this on linux, http://www.ibh-dd.de/~beck/stuff/lzs4i4l/ apparently the only thing it doesn't do is ascend's original predecessor of rfc1974 but it seems at least the later ascends can also be configured to do `proper' rfc1974, they only (well...) do the older thing by default. (ok there also seem to be patent issues with this type of compression and that linux source is GPL'd too so it probably at best could become a port, but that shouldn't stop it from being useful.) And the last thing, is anyone working on moving more of ppp back into the kernel, like, by using netgraph? (i hadn't really looked at this netgraph thing yet until i read the daemonnews article today... impressive stuff.) and is someone working on linking i4b and netgraph? that seems to be the logical way to do more complex stuff like this aodi thing that e.g. the german Telekom wants to use for their low-bandwidth 10 DEM/month isdn `flatrate' which they plan to introduce around the end of the year. (and _if_ this really works it sure will become pretty popular over here as long as all the other `real' flatrates are still in the 100 DEM or more range... :/ ) this seems to be the current draft: http://www.ietf.cnri.reston.va.us/internet-drafts/draft-ietf-pppext-aodi-02.txt Regards, -- Juergen Lock <nox.foo@jelal.kn-bremen.de> (remove dot foo from address to reply) To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200003042243.XAA82879>