Hello,
 
With similar setup I have experienced the same behavior - the only difference
was that I had another NIC plugged into router (eth2) and was bridging theese
two interfaces ( eth0 & eth2 ) while eth1 was connected to the gateway.
 
So: bridging enabled & proxy_arp on all interfaces enabled. I blamed the bridge
for the broblems ( since it was marked as experimental code ) and with kernel
without compiled bridge support, It worked flawlessly. Otherwise, the connections
were randomly shot down. And the computers were unable to connect again
for a while... strange. If I understand how bridge and proxy_arp works, there
is a big conflict here, unless everything is setup correctly ( ie - proxy_arp
on one interface, bridging only the second and third )
 
But... this may not be related to your problem.
 
I think that the problem is related to proxy_arp on older kernels like 2.2.13.
( Actually if I remember correctly, the problem was that the ARP response
was sent through the different interface that did not receive the ARP request )
 
Not sure whether this will help, but try proxy-arping only eth1 by setting this:
/proc/sys/net/ipv4/conf/eth1/proxy_arp = 1
 
( and all computers on your internal network should have default-gw set to your
router now, however, it is easy to set up if you are using bootp + dhcp )
 
Or: Upgrade to newer stable kernel. It might help...
Regards,
Petr
 
--
Petr Sebor
Technology/Engine Development
SCS Software [ http://www.scssoft.com ]
PGP fingerprint: 2412 1AEA 9673 8A5C  3C9F EF55 86B6 A03E  EA97 ED1D
 
 
----- Original Message -----
Sent: Wednesday, April 05, 2000 11:14 AM
Subject: Strange routing problem

Hello,

I have a transparent Linux firewall, now running slackware with kernel
2.2.13.

The firewall works very nice, and i get quite high transferrates, but
sometimes, almost at random, the firewall stops routing, one or all of
the machines looses Internet connection, and cannot ping any machines
on the other side.

Im quite bothered of this since i cannot find any fault, and i cannot
find any answer of why just some machines looses route sometimes, and
other machines another time. The loss is from 30 minutes to 2 hours,
and after that everything is normal, even when i was not touching any
of the machines.

The Linux firewall can ping all hosts inside and outside, and if i use
iptraf or an other stat program, it will show that 0 tcp packets is
send or received.

The setup is
/proc/sys/net/ipv4/ip_forward = 1
/proc/sys/net/ipv4/conf/all/proxy_arp = 1

my gateway is 2xx.xx.xx.129
eth0 nic is 2xx.xx.xx.156 (internal)
eth1 nic is 2xx.xx.xx.130 (external)
and the machines internal is from 2xx.xx.xx.131 to 2xx.xx.xx.155
well actualy .128 to .158 or something.

and the route is:
2xx.xx.xx.129   0.0.0.0         255.255.255.255 UH    0      0        0 eth1
2xx.xx.xx.128   0.0.0.0         255.255.255.224 U     0      0        0 eth0
127.0.0.0       0.0.0.0         255.0.0.0       U     0      0        0 lo
0.0.0.0         2xx.xx.xx.129   0.0.0.0         UG    0      0        0 eth1

--
Best regards,
 Fredrik Bergstrom


-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]

Reply via email to