I tryed using the isp's dns and stoping named so there wouldn't be all
of that dns traffic, and started to ftp a file from ftp.netbsd.org and
the fastest throughput I saw was 0.8 K and usually it was around 0.76
K and stopped it
I just reallized the message I sent with the setup files didn't go to
the list so I sent it a little while ago
I did change the line
115200
to
speed 115200

12/28/05, Mike M. Volokhov <mishka (AT) intostroy (DOT) comwrote:
Tue, 27 Dec 2005 20:53:27 -0600
Rick McCombs <rick.mccombs (AT) gmail (DOT) comwrote:

I tryed using the isp's dns and stoping named so there wouldn't be all
of that dns traffic, and started to ftp a file from ftp.netbsd.org and
the fastest throughput I saw was 0.8 K and usually it was around 0.76
K and stopped it

Hi!

What a speed was achieved under Slackware? Well, using NetBSD pppd will
not guaranteed that you will have hi-speed connectivity with NetBSD.org
site ;-) Please try to use the closest ftp/www servers to test your
connectivity. Your ISP should be good here.

With Slackware around 4K. In fact on the Win XP box with the sorry
winmodem it has I can get 4K from netbsd.org.

My ISP doesn't seem to have an ftp server an using traceroute I'm
unable to find anything close to me netwise.
Also, you may wish to run "pppstats -w 1" to see real-time ppp
statistics. A traceroute(8) and ping(8) utilities might be also helpful
to find narrowings.

I try that.

12/28/05, Mike M. Volokhov <mishka (AT) intostroy (DOT) comwrote:

Also, you may wish to run "pppstats -w 1" to see real-time ppp
statistics. A traceroute(8) and ping(8) utilities might be also helpful
to find narrowings.

pings and traceroutes seem about normal.
File transfers stall frequently and never reach 0.9 K bytes.
It's like watching grass grow.

In article <@mail.gmail.com>,
Rick McCombs <rick.mccombs (AT) gmail (DOT) comwrote:
12/28/05, Mike M. Volokhov <mishka (AT) intostroy (DOT) comwrote:
>>
>Also, you may wish to run "pppstats -w 1" to see real-time ppp
>statistics. A traceroute(8) and ping(8) utilities might be also helpful
>to find narrowings.
>
>pings and traceroutes seem about normal.
>File transfers stall frequently and never reach 0.9 K bytes.
>It's like watching grass grow.

I get 2.0 - 2.9 Kbytes with my:

com0 at pcmcia0 function 1: <3Com Corporation, 3C562D/3C563D, EtherLink III, LAN+Modem PC Card>

I see frequent stalls too, but the average transmission rate seems fine.
I do set my mtu to 296 (since I am not using ipv6).

christos

12/28/05, Christos Zoulas <christos (AT) astron (DOT) comwrote:

I get 2.0 - 2.9 Kbytes with my:

com0 at pcmcia0 function 1: <3Com Corporation, 3C562D/3C563D, EtherLink III, LAN+Modem PC Card>

I see frequent stalls too, but the average transmission rate seems fine.
I do set my mtu to 296 (since I am not using ipv6).

Does that really help? I heard people swear by different mtu settings
and I couldn't tell the difference, but I haven't tried in a while.
However isn't it the mru that would matter for downloads?

29 Dec 2005 09:00:45 -0500, Greg Troxel <gdt (AT) ir (DOT) bbn.comwrote:
I also use ppp, on current from September with a USR sportster 33.6
external, dialing the same kind of modem on another NetBSD machine.
Throughput/speed seems to be fine, limited by the modem.

I do have another problem, which is that when the line goes down (due
to modem hangup - this is a nailed-up connection), pppd notes this,
unconfigures the interface, and then hangs in a wedged state where it
doesn't exit, or worse, do the redial commanded by persist.
But I think this is unrelated to performance issues.
I got past the problem I was having but

I discoved the same problem, that persist doesn't work. My ISP
disconnect me after 8 hours no matter what.
(I would change but my dad is paying for it and they are about the
cheapest. It used to be my mother's account and she died and my dad
has kept paying it so I could sell stuff.)

I had persist in my options and was having it buld some stuff in
pkgsrc and after the modem disconnected it didn't ever reconnect and
there was several pppd processes.