Date: Tue, 23 Jan 1996 08:22:19 -0800 From: gshaffer@nosc.mil (Greg Shaffer) To: current@freebsd.org Cc: gshaffer@nosc.mil Subject: Re: finger problem going from 2.1R to -current Message-ID: <v02140b01ad2ab5348f6d@[128.49.16.48]>
next in thread | raw e-mail | index | archive | help
I have been following this thread for awhile, no let me throw some fuel on
the fire!
I've had some problems over the past several weeks with what appeared to be
some routing problems, but I now believe they are related to the T/TCP
(more specificly RFC 1644) implementation in FreeBSD 2.1R. To help
illustrate my problem here is a simple network diagram.
A B
+-------+--------+
| (Ethernet)
(FW) |
+--+-------C
| |
| | (PPP)
| +-------D
|
|
+-------E
Where: Machine C, D and E are FreeBSD 2.1R.
Machine A is SunOS.
Machine B is Solaris 2.5.
Machine A, B, C and D are all within the same domain
behind a firewall.
Machine E is outside on the Internet.
Here is the problem.
>From machine D I could ping and traceroute to machine C and E. I could not
use and other services (e.g. ftp, rlogin, telnet, http, etc). I would have
to go through and intermediate machine (e.g. A or B) to get to C. When ever
I tried to access one of these machines (C or E) from D the connection
appeared to startup up and then hang. Netstat showed that the connection
had been made but output at the terminal never completed (i.e. I never saw
the login prompt). For a while I thought I had a routing problem, but when
I boot machine D using Windows I could get to machine C and E with no
problems. When I first read this thread I felt the problem might be
remotely related. When I disabled RFC 1644 support on machine C and D I was
able to get into both C and E from D without any problems.
Forgive me if this sounds like a silly question, but did anybody test T/TCP
between two FreeBSD 2.1R machines?
Greg Shaffer
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?v02140b01ad2ab5348f6d>