On Thursday 05 June 2003 10:49, giangy wrote:
Do you have a routing table entry for net 10.30.3.x? This doesn't look like a bug to me. What do the rule counters say? What about tcpdump?
Bart
IT WORKS!!!
adding routing table entries for remote nets now redirect works!
Since I could reach remote hosts WITHOUT routing entries... why does the kernel need it???
This is just a guess, since I haven't tried it, but here it goes...
In your configuration without a direct route to 10.30.3.x:
When you ping from the bridge, the packet is sent to your default gateway and then to the host.
No.... I traced it ... it goes directly to the host.
Bridge finds out (at first time 10.30.3.x sends a packet out to internet) where the host is and which gateway to use (actually I was really impressed how well the birdge works)
It was just curiosity.... since it works ... it's fine with me
thanks again
gianluca
When the 10.30.3.x host sends a packet that is redirected on the bridge, the bridge wants to reply directly to the 10.30.3.x host and doesn't have a direct route, so the bridge drops the packet.
cheers, Bart
_______________________________________________ Bridge mailing list [EMAIL PROTECTED] http://www.math.leidenuniv.nl/mailman/listinfo/bridge
