home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Columbia Kermit
/
kermit.zip
/
newsgroups
/
misc.19950726-19950929
/
000364_news@columbia.edu_Fri Sep 15 13:51:56 1995.msg
< prev
next >
Wrap
Internet Message Format
|
2020-01-01
|
5KB
Received: from apakabar.cc.columbia.edu by watsun.cc.columbia.edu with SMTP id AA01787
(5.65c+CU/IDA-1.4.4/HLK for <kermit.misc@watsun.cc.columbia.edu>); Fri, 15 Sep 1995 09:51:59 -0400
Received: by apakabar.cc.columbia.edu id AA18423
(5.65c+CU/IDA-1.4.4/HLK for kermit.misc@watsun); Fri, 15 Sep 1995 09:51:57 -0400
Path: news.columbia.edu!merhaba.cc.columbia.edu!chaiklin
From: chaiklin@columbia.edu (Seth Chaiklin)
Newsgroups: comp.protocols.kermit.misc
Subject: Progress on MSK 3.14 and Linux 1.2.8 problem
Date: 15 Sep 1995 13:51:56 GMT
Organization: Columbia University
Lines: 97
Message-Id: <43c0ds$hvk@apakabar.cc.columbia.edu>
Nntp-Posting-Host: merhaba.cc.columbia.edu
Apparently-To: kermit.misc@watsun.cc.columbia.edu
This message is a continuation of my inquest into why I have
had problems with MSK 3.14 and Linux connections.
I have made some progress, I believe in delimiting the problem,
so I will start a fresh thread.
I posted a question about these problems to the Linux newgroups and
after several days, I got the following reply which I send further
without permission (because it was publically posted). The main
gist of the message is that it is believed that Linux ARP is OK.
Here it is:
------------------------
Newsgroups: comp.os.linux.networking,comp.os.linux.help,comp.os.linux.misc
From: iialan@iifeak.swan.ac.uk (Alan Cox)
Subject: Re: ARP cache problem in 1.2.*??
Organization: Institute For Industrial Information Technology
Date: Wed, 13 Sep 1995 13:29:07 GMT
In article <433jha$k6m@apakabar.cc.columbia.edu> chaiklin@merhaba.cc.columbia.edu (Seth Chaiklin) writes:
>I have also tried MS-Kermit Version 3.14, 24 May 1995, and that will
>simply not connect to the Linux box, complaining about Arp resolution
>problems. I have communicated with the Kermit people aboout the
>problem and I am still trying to figure out what is going on, but
>one candidate is that the Linux ARP cache is not well-behaved yet.
Linux ARP is believed to work, and its a 'seen by nobody else' problem.
That given the userbase normally means its a user setup problem. The one
person here who persists in using kermit under DOS, his own custom mail
collect read/write scripts and other strange things hasn't moaned at me,
nor have other kermit + Linux users I know.
>I have tried to look in the kernel changes and could see that some
>work was done on Arp. Can someone comment on whether it is likely
>that the problems with Kermit could be located in the Linux ARP cache?
The only ARP changes made recently have been in the 1.1.x kernels to clean
up code and add netmasks to proxy arp (for transparent routing) and some
recent 1.3.x fixes whereby ATF_PERM entries would previously be updated
by an ARP reply for that address.
---------------------------------------------
Meanwhile, this mention of apparent success by another person made
me wonder what will happen if I used another version of MSK. So I
make the following experiment.
1. Run MSK 3.14 on a 486 machine with ODIPKT and 3C509 card.
(and no energy saving features)
As usual, after 10 - 15 minutes, the ARP entry disappears,
the screen output disappears, though it is still possible
to enter commands.
2. Powerdown machine.
3. Run MSK 3.13pl5 on the same machine.
a. After 10 minutes, the entire IP number (as opposed to HW address)
disappears from ARP table.
b. Press CR on MSK. It works.
c. Let sit 45 minutes.
d. Press CR. It still works.
e. Logout, manually delete IP number from ARP cache with arp -d
f. Powerdown MSK machine. Login again with MSK 3.13
g. Wait 45 minutes. At this point the HW address has disappeared
from the ARP cache (i.e., 00:00:00:00:00:00), even though the
IP number is still there. It still works!
h. Delete IP number (arp -d) while MSK 3.13 is connected. It still
works.
i. Login with MSK 3.14 (18 Jan 95) edition.
j. Delete IP number from ARP cache (arp -d). It freezes!
k. If I add and delete the IP number by hand a few times, I can
"jumpstart" the MSK 3.14 machine back to monitor output.
I am willing to suggest that there is some change between MSK 3.13 and
MSK 3.14 (18 Jan 95) that is critical to this problem. And there is
yet another change from MSK 3.14 (18 Jan 95) to MSK 3.14 (24 May 95) that
creates a new problem.
Well, I don't know much more now about why these problems occur than
I knew yesterday, but I am starting to get a better sense of how to
create them.
Cheers,
Seth Chaiklin