Hi, is it anyhow possible to tell the other end of the ppp link how he should queue the packages? I have a ppp link to my isp and would like to control the order of packages send to me. I would like to give uucp a lower priority than http. Is this possible? Has ppp any special features to control such things? Bye, Jörg. -- Gott hat den Menschen erschaffen, weil er vom Affen enttäuscht war. Danach hat er auf weitere Experimente verzichtet. (Mark Twain)
I would say no. You could only shape (= drop) some of the incoming packets, and hope that the sender will slow down on this, because he doesn''t receive ACK for the dropped packets. This you can do with the normal tc methods. Andreas Joerg Sommer (joerg@alea.gnuu.de) schrieb:> > Hi, > > is it anyhow possible to tell the other end of the ppp link how he should > queue the packages? I have a ppp link to my isp and would like to control > the order of packages send to me. I would like to give uucp a lower > priority than http. Is this possible? Has ppp any special features to > control such things? > > Bye, Jörg. > -- > Gott hat den Menschen erschaffen, weil er vom Affen enttäuscht war. > Danach hat er auf weitere Experimente verzichtet. > (Mark Twain) > _______________________________________________ > LARTC mailing list > LARTC@mailman.ds9a.nl > http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc >
On Thursday 13 October 2005 10:52, Joerg Sommer wrote:> is it anyhow possible to tell the other end of the ppp link how he > should queue the packages?Only by making a phone call and threatening the ISP directly... ;-)> I would like to give uucp a lower priority than http. Is this possible? > Has ppp any special features to control such things?It''s possible, although there is not really a nice solution for it. I do this kind of "prioritizing" by shaping packets that the router can send to the machines in my LAN, which basically reduces to "drop anything that comes in too fast", in your case "drop uucp packets if they come in too fast". Not really nice since packets have to be re-sent, but for me this is still better than letting one file transfer hog the whole line forever when I need bandwidth for interactive sessions. Regards, Andreas Klauer
Apparently Analagous Threads
- Bug in 2.6.0-9
- 6398779 Need to reduce address limits for devices downstream of a BCM 5708/5714/5715 bridge
- No upstreams on domU, but downstream works fine
- Is that an efficient way to find the overlapped , upstream and downstream rangess for a bunch of rangess
- LLVM Releases: Upstream vs. Downstream / Distros