Date: Tue, 26 Oct 1999 17:04:05 -0700 From: "Craig Critchley" <craigc@nwlink.com> To: "Garrett Wollman" <wollman@khavrinen.lcs.mit.edu>, <net@FreeBSD.ORG> Subject: Re: FTP Net Performance Message-ID: <1f6601bf200e$c8e476b0$0201010a@fuzzer.com> References: <1df701bf1f84$32751020$0201010a@fuzzer.com> <199910261606.MAA83232@khavrinen.lcs.mit.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
Thanks. I guess I need to track down the thesis and give it a read - the
graphs themselves are relatively uninteresting - flat lines at large
negative numbers. (This is a small enough network that I can make sure
there is more or less nothing else happening when I try to FTP something.)
I got conflicting suggestions from a couple people about whether to use full
or half-duplex. The device defaults to half-duplexing, which seems to be
correct for a hub (the two machines do not have the switch between them as
they are both on the 100Mb segment). I'm not too concerned about 50Mb vs.
100Mb, or even 10Mb vs. 100Mb - I'm getting barely 100kilobits. Three
orders of magnitude difference. But uploading to the BSD machine goes full
speed. (For the record, ifconfig and the hub blinkies agree that I have a
100Mb link.)
Looking at netstat -i, I see that I'm getting Oerrs of about 15% of Opkts,
but no Ierrs. There is an inconsequential number of collisions. Perhaps
the error rate is to blame for the problem. What sorts of things do these
errors represent?
Thanks,
...Craig
----- Original Message -----
From: Garrett Wollman <wollman@khavrinen.lcs.mit.edu>
To: Craig Critchley <craigc@nwlink.com>
Cc: <net@FreeBSD.ORG>
Sent: Tuesday, October 26, 1999 9:06 AM
Subject: FTP Net Performance
> <<On Tue, 26 Oct 1999 00:32:02 -0700, "Craig Critchley"
<craigc@nwlink.com> said:
>
> > Any suggestions profoundly appreciated.
>
> Whenever TCP behaves strangely, your first step should be to capture a
> trace with tcpdump(8) and use tcpdump2xplot.pl and xplot(1) (part of
> /usr/ports/math/xplot) to see what's going on. The visual analysis
> technique is developed in Tim Shepard's S.M. thesis, references to
> which are available in the xplot tarball.
>
> -GAWollman
>
> --
> Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the
same
> wollman@lcs.mit.edu | O Siem / The fires of freedom
> Opinions not those of| Dance in the burning flame
> MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad
Irschick
>
>
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-net" in the body of the message
>
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1f6601bf200e$c8e476b0$0201010a>