You gurus out there: What are the advantages/disadvantages of bridging vs. routing with wireless interfaces?
Is there more cpu overhead with bridging two wireless channels than with simply routing them? Intuitively, it would appear that bridging has a higher overhead since it has to maintain two interfaces and then join those together into a new br0 interface and maintain all of that in software before any packet routing takes place. I am using a two-channel Soekris board (486-133mhz, 64mb ram, 16mb CF) with two 200mW Teletronics (Senao) pcmcia radios with both running in A.P./Master Mode. Attaching 120-degree sector antennas (North and West) to the two cards, which would be the more efficient: 1. Assign each interface/antenna to its own subnet, i.e., netcs0=192.168.2.0/24 and netcs1=192.168.3.0/24, with separate ESSID for each sector (WISP_N, WISP_W) and use dhcpd to hand out addresses on each interface? (This is my current setup) 2. Bridge the two into one interface, br0; setup dhcpd to lease addresses on only one subnet? (I haven't had any luck getting the two to bridge using Wisp-Dist, but I assume that it IS possible.) 3. Another option I have not considered? With the limited amount of cpu power in the 4521 it makes sense to find the most cpu-efficient way of handling high packet loads. The additional complexity of maintaining two subnets would be justified if doing so lowers the overhead on the cpu. CPU loading becomes especially critical when WEP encryption and iptables firewall filtering loads are added. Dan ------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 _______________________________________________ leaf-wisp-dist mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-wisp-dist