home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Columbia Kermit
/
kermit.zip
/
newsgroups
/
misc.19941031-19941221
/
000233_news@columbia.edu_Tue Nov 22 14:31:55 1994.msg
< prev
next >
Wrap
Internet Message Format
|
2020-01-01
|
3KB
Received: from apakabar.cc.columbia.edu by watsun.cc.columbia.edu with SMTP id AA11635
(5.65c+CU/IDA-1.4.4/HLK for <kermit.misc@watsun.cc.columbia.edu>); Tue, 22 Nov 1994 09:32:01 -0500
Received: by apakabar.cc.columbia.edu id AA27123
(5.65c+CU/IDA-1.4.4/HLK for kermit.misc@watsun); Tue, 22 Nov 1994 09:31:59 -0500
Path: news.columbia.edu!watsun.cc.columbia.edu!fdc
From: fdc@watsun.cc.columbia.edu (Frank da Cruz)
Newsgroups: comp.protocols.kermit.misc
Subject: Re: Help: file transfer problem
Date: 22 Nov 1994 14:31:55 GMT
Organization: Columbia University
Lines: 30
Message-Id: <3asvcr$qfg@apakabar.cc.columbia.edu>
References: <CzMLA2.K01@liverpool.ac.uk> <3aqlhu$7q6@apakabar.cc.columbia.edu> <CHANG.94Nov21182356@theta.math.wsu.edu>
Nntp-Posting-Host: watsun.cc.columbia.edu
Apparently-To: kermit.misc@watsun.cc.columbia.edu
In article <CHANG.94Nov21182356@theta.math.wsu.edu>,
Ching Mo Chang <chang@theta.math.wsu.edu> wrote:
>I also have the file upload problem with MS-Kermit (Also C-Kermit in OS/2),
>I can download file with packets length at 5012, but when I use the same
>setting to upload file I got lots of timeout and eventually end up too many
>retries error and fail to upload file. If I really need to upload file, I
>need to increase the retry limit and set the packets length to 92, let the
>file crawl from my PC to the host or I will just paste the file into an editor
>in my host side (only work in ASCII file).
>
This is not Kermit's fault. You should be thankful that Kermit lets you
adjust these parameters to make the file transfer work at all. Why is this
happening? It's hard to say without more information, but the most likely
culprit is a lack of buffer capacity in the "upstream" direction, coupled
with a lack of effective flow control -- a fatal combination, and a common
one.
Many communications processors (terminal servers, front ends, host console
drivers) are designed on the assumption that traffic *to* the host consists
of nothing but keystrokes -- which hardly anybody can produce at more than
about ten per second (= 100 bps) -- whereas traffic in the downstream
direction is voluminous -- file listings, etc. So they have big output
buffers and tiny input buffers.
To compound the problem, some communications processors take this assumption
one step further and do not even provide flow control in the upstream
direction, because they figure they will never need it. A well-known example
is the Cisco ASM series of terminal servers.
- Frank