Dead Gateway Detection with PPPoE

Linux Advanced Routing and Traffic Control

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Gentlefolk,

	First, many thanks to EVERYONE that tries so hard to make this advanced 
routing stuff useful to the "...rest of us"! You all rock!
	I have been prowling the archives of this list for an answer to my 
problem, and have seen some close situations, but no joy. Yet.
	I've got a relatively simple setup I'm trying to get working: we've got 
a public-access network with 90 or so clients and I'd like to route all their 
traffic over a 3 Mbps DSL line - except when that goes down, then I'd like to 
fall over to our T1.
	So I'm working on a Debian box in hopes that it will act as a router for 
this public subnet like so:

ppp0	DSL IP 		(via PPPoe)
eth1	192.168.1.0	(Public subnet)
eth2	XX.XX.XX.XX     (Our DMZ)

	I've downloaded the 2.4.31 kernel source, patched it with the 
Julianpatches and read nano.txt about 8 times.
	I am SNATing the outgoing traffic to the appropriate address with 
iptables depending on the interface like so,

iptables -t nat -A POSTROUTING -s $NATNET -o ppp0 -j SNAT --to $DSL
iptables -t nat -A POSTROUTING -s $NATNET -o eth2 -j SNAT --to $WRL

	I made sure to include the stateful iptables statements as they appear 
in nano.txt.
	I worked extensively with mpath2.sh and have since cobbled together a 
script a la nano.txt, with the associated ping script to test routes by pinging 
each of the two gateways every 60 seconds.
	When initially running the script it sets up the tables just the way one 
would expect and everything works great. Clients are happy and can access 
the web at blazing speed. When the DSL line goes down, however, 
the routing table doesn't change to reflect the T1 gateway and the clients are 
dead in the water - no pinging of hosts not directly connected to the router box 
and no web traffic even though domain names are resolving. 
	When the DSL line comes back up that route is used instantly and the
clients are all happy again.
	So I'm clearly missing something.
	I include some info below, and I'm happy to get more specific upon 
request, I just don't want to flood the list. 
	Can anyone suggest other things I can check?
	I'm grateful for any help anyone can take the time to offer!


root@endurance:~# ip rule
0:      from all lookup local 
50:     from all lookup main 
201:    from (Correct DSL IP address) lookup 201 
202:    from (Correct DMZ network) lookup 202 
222:    from all lookup 222 
32766:  from all lookup main 
32767:  from all lookup default 


(When DSL is UP)
root@endurance:~# ip route get 204.152.189.113
204.152.189.113 via (DSL gateway) dev ppp0  src (DSL IP address)
    cache  mtu 1492 advmss 1452


(When DSL is DOWN)
root@endurance:~# ip route get 204.152.189.113
RTNETLINK answers: Network is unreachable
-- 
********************************************************************
Brett Charbeneau
Network Administrator         
Williamsburg Regional Library                   
7770 Croaker Road                               
Williamsburg, VA 23188-7064                     
(757)259-4044          www.wrl.org
(757)259-4079 (fax)    brett at wrl.org
********************************************************************



_______________________________________________
LARTC mailing list
LARTC@xxxxxxxxxxxxxxx
http://mailman.ds9a.nl/cgi-bin/mailman/listinfo/lartc

[Index of Archives]     [LARTC Home Page]     [Netfilter]     [Netfilter Development]     [Network Development]     [Bugtraq]     [GCC Help]     [Yosemite News]     [Linux Kernel]     [Fedora Users]
  Powered by Linux