Just a quick note -- I describe my network as featuring a Linux
packet-filtering router, which is not entirely accurate. Iptables was not
active during bridging experiments. Well, iptables may have been active,
but all default chain policies were set to ACCEPT, and all rules were
flushed.
Eric
"Eric Adamson" <(E-Mail Removed)> wrote in message
news:8WO6e.6441$_(E-Mail Removed)...
> My home network consists of a Linux packet-filtering router serving a
> handful of Win2K & WinXP boxes and two laptops (not diagrammed below),
> connected within two Class C subnets: 192.168.1.0\24 (802.11b) &
> 192.168.2.0\24 (10BT).
>
> Here's a simple network diagram: (view using Courier font or equiv.)
>
> wlan0 eth0
>
> +---------+ +---------+------+ +-------+-------+
> | | | | | | | |
> cable_modem wireless_router box_1 Linux_router box_2 printer
>
> I would like to bridge wlan0 & eth0, merging the two subnets. All
addresses
> on wlan0 & eth0 are static -- the wireless router is instructed to
recognize
> machines by MAC ID. I've loaded the newer bridge module (bridge.o) per
the
> HOWTO, and hosts on the two subnets could see each other, but box_2 could
> not see the wireless_router.
>
> Assuming this configuration is workable, how should routing work: should
> box_2's default gateway be the wireless router's IP, or that of the Linux
> router? I would expect it to be the wireless router, but box_2 cannot see
> it. The bridge doesn't appear to be completely transparent -- do I need
> help from iptables? Please note that the Linux router will also be a host
> on the merged network -- does this make a difference?
>
> I suspect I haven't given enough information, but this is all I have time
to
> write, for the moment. Any ideas or suggestions would be greatly
> appreciated.
>
> Thanks,
>
> Eric Adamson
> Lansing, Michigan
>
>
|