[e2e] Re: crippled Internet

Vernon Schryver vjs at calcite.rhyolite.com
Fri Apr 27 08:16:30 PDT 2001

> From: "David P. Reed" <dpreed at reed.com>

>                                          ...  they don't realize the delays 
> introduced by V.92 compression and PPP are serious problems.  I haven't 
> measured the delay & jitter introduced by PPPoEt (used by almost all DSL 
> broadband providers, and possibly some cable modem providers), but PPP may 
> well be problematic there as well, if there is competing traffic. ...

PPP is not v.92, v.42, nor v.42bis and is not like any of them.  Async
modem compression and packetizing such as v.42 and v.42bis require
the modem to wait to accumulate a packet of data, but PPP is purely
sync.   (yes, the modem DCE-DCE links are sync, but the DTE-DCE links
are usally async.).  V.42 also differs from PPP by delaying data to
retransmitt after errors.

A reasonable PPP implementation introduces no significant delay other
than that forced by congestion.  That obvious when you stop to think
that at least some big routers use PPP to talk to each other.  Even
when CCP (payload) or VJ Header compression are used, PPP still does
not delay any packet by more than the sum of the serialization delays
of the packets ahead in the queue.

In my view, a reasonable PPP implementation honors the IP TOS bits as
well as optionally watching for interesting UDP and TCP port number
and might even move small packets to the front of the transmit queue,
and so gives low-latency traffic delays as good as possible.

On the other hand, if by "PPP delay", the idea was not to confuse PPP
with v.92 or v.42bis but to talk about ordinary and unavoidable queue
and serialization delay, or if the idea was to point out that for bulk
data on modems, it is best to run PPP with both v.42 and v.42bis,
then never mind.

(Bulk data on a modem wants link layer compression like v.42bis because
STAC CCP is lame and even it is often not supported by access routers.
It wants link layer retransmissions like v.42 because modem links are
often too noisey for PPP's drop-on-FCS-failure tactic.)

Vernon Schryver    vjs at rhyolite.com

More information about the end2end-interest mailing list