[pptp-server] Re: [LRP] cant ping the client and others

Christopher Kuhl chrisk at ciris.net
Thu Nov 4 19:08:53 CST 1999

Actually, I am connecting to an isp an each side, one with isdn, and the
other just an analog modem.  On the server side I start up popotop and wait
for incoming GRE connections.  On the client side I start a GRE session with
the linux pptp client.  The gre session is between the ip's that I get from
my isp's.  Once this tunnel is up, my poptop server starts a ppp session
through the tunnel.  This is how I can use the 10.111.XXX.XXX ip's, they are
being encapsulated accross the net, so they are not vissible during
transport.  This is basicaly how vpn's work.  The remote ip's I was
referring to is the ip's that pppd assigns and passes to the ip-up and
ip-down scripts.  One the client side, I get a local and remote ip, the
remote is my gateway back to the host poptop server, and the same goes for
server side.  The ip that it issues me is it's gateway back to my network.
I am now considering running ciped on another port and use there ipip
solution to see if I still have the same routing problems with another
tunnel, and just use poptop for windows clients, which work fine right know.

Thanks for the input, and let me know if you think of something.
-----Original Message-----
From: Ray Olszewski <ray at comarre.com>
To: Christopher Kuhl <chrisk at ciris.net>; linux-router at linuxrouter.org
<linux-router at linuxrouter.org>
Date: Thursday, November 04, 1999 6:15 PM
Subject: Re: [LRP] cant ping the client and others

>Since you've posted this question several times without getting a response,
>I assume there aren't any experts on this stuff hanging out on this list.
>I'm no expert, but my next project will be to implement one of the VPN
>technologies, so I took a little time to look over your setup. Bearing in
>mind my limited expertise here, I have one thought ...
>How are these two routers actually connected? You show them as connected
>through "internet", which I intrepret to mean that each connects to an ISP,
>and the ISPs find a route between the two LANs. Under that assumption, what
>is the "remote ip that is passed by pppd" address you are giving to the
>client? I would normally think of this phrase as referring to the ISP's end
>of the PPP connection to the client. If that is what you mean, why would
>that host know how to route to your private subnet-10 network on the other
>end? I would think that each of the routers needs the ppp address of the
>other router here (or they need to be masq'd, which you say works, at least
>after a fsahion).
>Apologies in advance if I've misinterpreted your posting. I'll be happy to
>take a closer look at your configuration if you're interested, but pleas
>understand that I'm offering because I need to figure this stuff out too,
>not because I already have any great expertise in it. Good luck in any
>At 02:07 PM 11/4/99 -0600, Christopher Kuhl wrote [in part]:
>My setup is this:
>>PTP client( network(
>>I am using poptop on one side and the linux pptp client on the other.
>>I connect to the Poptop server with a Windows machine I can ping and use
>>entire network fine.  But from the linux pptp client, I can only ping the
>>Poptop router(ping and vice versa.  I can not ping other
>>machines on the remote network.  I changed my ip-up and down scripts to
>>and delete routes like this
>>Poptop side:
>>route add -net netmask gw $5 $1   ;  $5 is the
>>address I am giving the client router.
>>Cleint side:
>>route add -net netmask gw $5 $1   ; $5 is the
>>remote ip that is passed by pppd.
>>I checked my routing tables, and they are correct.  I also delete these
>>routes when ip-down runs.
>[rest deleted]
>------------------------------------"Never tell me the odds!"---
>Ray Olszewski                                        -- Han Solo
>Palo Alto, CA                    ray at comarre.com

More information about the pptp-server mailing list