From owner-freebsd-hackers  Sat Nov  4 15:34:05 1995
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id PAA23312
          for hackers-outgoing; Sat, 4 Nov 1995 15:34:05 -0800
Received: from scam.XCF.Berkeley.EDU (scam.XCF.Berkeley.EDU [128.32.43.201])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id PAA23299
          for <freebsd-hackers@freebsd.org>; Sat, 4 Nov 1995 15:34:03 -0800
From: grady@scam.XCF.Berkeley.EDU
Received: from localhost (localhost [127.0.0.1]) by scam.XCF.Berkeley.EDU (8.7/8.7) with SMTP id PAA12852 for <freebsd-hackers@freebsd.org>; Sat, 4 Nov 1995 15:34:23 -0800
Message-Id: <199511042334.PAA12852@scam.XCF.Berkeley.EDU>
X-Authentication-Warning: scam.XCF.Berkeley.EDU: Host localhost [127.0.0.1] didn't use HELO protocol
To: freebsd-hackers@freebsd.org
Subject: PPP problems with ISDN "modem"?
Date: Sat, 04 Nov 95 15:34:23 -0800
Sender: owner-hackers@freebsd.org
Precedence: bulk

I've hit a few problems using ppp (iijpp v. 0.94) under FreeBSD 2.0.5.
The first one I hit was that I was getting the message:
	** Too many ECHO packets are lost. **
after about three minutes of connection time, fairly consistently.
I was also getting this with my 28.8 connection, but it would only happen
every few hours there.  I've put what appears to be a short-term fix
for this by setting the LQR timeout to be one day (i.e.:
    set timeout 300 86400
).  I'm not sure yet whether this is a long-term fix.

A more serious problem right now is that most large-file transfers
are hanging after a small percentage of the file is transferred.  It
seems to be due to an interaction with the IP software on the two
machines.  For instance, one file I was transferring (ppp.tar.gz)
would always hang after 18980 bytes had been transferred, when transferred
from a SunOS 5.4 machine, but when I put the file onto a different machine
(running Ultrix v4.3a), it transferred fine the first time.  From the sun,
it would hang after around 24k in ASCII mode.  While it was hanging
(in ESTABLISHED mode), other connections (e.g. rlogin) continued to
work fine.

Similarly, while using Netscape (1.1), most pictures would hang during their
transfer (although text usually managed to transfer easily).

Finally, I just seem to have hit the rough edges of the ppp software
in general. For the first time, I got a core dump from ppp (with a trashed
stack).  I've seen problems in cleanup when exceptional situations (like
the "too many ECHO packets" problem above) cause the program to exit.
And, of course, as I try to use some of the more obscure features of
ppp, the lack of decent documentation has becomes more and more of a
problem.

Anyway, the main thing I wanted to find out was how to deal with the
problems with hanging.  Has anyone else seen this, and found a fix?

Thanks for any help.

	Steven