At 6/25/2010 10:45 PM, Rubens Kuhl wrote: > > OpenWRT and OLSR or BATMAN on a Routerboard or Ubiquiti CPU platform > > may be ideal, but I need to learn more about OLSR and BATMAN in > > practice. BATMAN seems to be a distance-vector algorithm, like, uh, > > DECNET 3 and 4 and IGRP, while OLSR is link state, like OSPF. I am > > partial to link state. The BATMAN guys note that it doesn't scale > > well, especially >100 nodes, but I'm not looking to have that many in > > a domain. Distance vectors are fast to learn new routes but have > > problems with dropped routes. > >Although TRILL is being developed on networks with fiber-rich diets, >it might be good to a wireless mesh: >http://en.wikipedia.org/wiki/TRILL_%28computing%29 > >In essence, Layer 2 link-state that is good for meshes. The question >if link-state or distance-vector is more appropriate to a wireless >mesh is something yet to be defined, but you said you are partial to >link-state, so TRILL will probably thrill you.
Yes, TRILL looks like a good idea. And since Radia Perlman wrote the RFC, I trust that it is of unusually high quality for an RFC. ;-) However, the only daemon I'm aware of is for OpenSolaris (Radia's at Sun), not a WRT. It sure would be nice if either of the dueling WRT teams implemented it. And btw I stand corrected on OLSR -- it's IP only -- BATMAN-adv is Layer 2. -- Fred Goldstein k1io fgoldstein "at" ionary.com ionary Consulting http://www.ionary.com/ +1 617 795 2701 -------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: [email protected] Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
