Modem Negotiated Speed

• Messages sorted by:[ date ][ thread ][ subject ][ author ]

Hi,

Russell King wrote:
> 
> Gert Doering writes:
> > Russell King wrote:
> > > I'm sorry, but I don't really see the purpose of logging the speed at all.
> > > Yes, I agree that there should be some way of calculating the server loads
> > > vs modem speeds, but the negociated speed is not representitive of the
> > > speed throughout the whole call.
> >
> > It's, for example, a good hint at "bad modem lines" -- if you get faster
> > averages on three lines than on the fourth, it smells of problems with
> > that line / that modem.
> 
> My personal experiance with "bad modem lines" does not reveal that this is
> always the case.  So far all of the "bad modem lines" suffer from noise that
> gets worse the longer that they are used.

I generally have six dialin users at a given time logged into my
server.  In a 24 hour period (say on Monday) the server disperses
(over 8 dialin modems total) an average of 150meg (Rx and Tx) per
modem.  Now, yes it is true transmission speeds change constantly. 
But even on a crystal clear connection you are *not* going to get 56
or 48 or even 28 through put when the dialin uses a 14.4 on thier
end.  Knowing what the modems connected with to each other eliminates
alot of assumption about bad modems and noisy lines.  As well as let
me know what my clientel is useing.  Knowing what they are capable of
helps me service them better.

You see :-)

> 
> > > Generally, if you have a really good line, you'll get about 56k through it.
> > You do? Never seen more than maybe 48000.
>