Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
(ASCEND) max 2000/linux stopped ppp connection
I'm experiencing a ppp connection stopping between the max 2000 we just
bought and my linux (red hat 5.1) P6 box at home, which came with ppp 2.3.3.
By `stopping' I mean that no packets flow between the two; the phone
lines are still connected, `show modems' says the interface is up, but I
cannot ping the max from home, and I cannot ping home from the max (once
I use another phone line to dial in through another system and telnet to
the max). (telnet, etc. don't work either, of course.) I can ping or
telnet anywhere else from the max, as expected, since that goes through
the ethernet.
This happens anywhere from a few seconds to an hour+ after dialing in.
(I've never had the connection stay good for two hours.) Sometimes it
happens when I'm ftp-ing a file, sometimes when I'm just messing about
in Emacs, whatever.
First I thought it was some kind of timeout. So I set idle=0 in
ethernet->connections->me->session (for now there is no radius, I'm the
only user, we're trying to set it up), and furthermore I'm pinging
through the connection to a far host every 90 seconds. (That ping also
fails when the connection stops, naturally.) I haven't been able to
find any other timeout parameters.
Then I thought the problem might be some incompatibility in the two ppp
implementations. I've turned off all the compression options I could
find, both at the max end (link comp=no, vj comp=no, etc.) and at the
linux end (nodeflate,noccp,novj,novjccomp,nopcomp,nobsdcomp). (The linux
syslog now reports that only <accomp> is being used.) Doesn't matter.
As for routing, the routes are still installed at both ends. There is
no `host unreachable' message, the pings simply get no reply. But in
case it matters, the basic scheme is that we have one subnet (call it
10.10.53) allocated for all the dialup hosts, so my home box gets
assigned, say, 10.10.53.101 (we're doing dynamic allocation from one
pool). The max ethernet interface has one address on the main network
to the rest of the world, call it 10.10.50, and a second address
(ethernet->mod config->ether options->2nd adres) on 53.
I've tried two different modems at home (a K56 internal one and a 33.6
USR sportster). Every call gets answered by a different line at the max
end. Never any difference in behavior.
Unfortunately, the latest ppp version (2.3.5) from
ftp://cs.anu.edu.au/pub/software/ppp/ppp-2.3.5.tar.gz doesn't compile on
my linux system, due to kernel changes, etc. I can pursue that, but I'm
skeptical that that is the problem.
I haven't yet tried connecting with anything other than linux ppp.
That's next. But I need linux ppp to work, in any case.
Any suggestions would be most appreciated. TIA.
karl@cs.harvard.edu
++ Ascend Users Mailing List ++
To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd: <http://www.nealis.net/ascend/faq>
Follow-Ups: