[Olsr-dev] mtu problem with multicast ipv6 packets
Mon Mar 23 20:48:16 CET 2009
On Montag 23 März 2009 20:15:45 you wrote:
> > Doing multicast and let the kernel handle the traffic (so that OLSR does
> > only need to initialize/update the multicast routes/trees).
> Yes but, what you want to do for creating the spanning tree ??
> Have you got already some smarter idea than PIM-DM or PIM-SM that
> takes advantage of OLSR information ?
My idea was to use some kind of statistics.
If the multicast group has only a few targets (3-4 ?) just unicast all data.
If it has at least 1/3 of the mesh network just flood it (with some addition
to skip parts of the net without targets at all).
If the number is between the two values, use a multicast tree based on the
routes to each target.
> At the University of Tor Vergata in here Rome we developed OBAMP
> https://netgroup.uniroma2.it/Andrea_Detti/obamp/ (theory)
> https://svn.ninux.org/ (code)
> it is an overlay multicast implementation that already works in Java.
> Overlay meaning that even not all the nodes in the mesh run the OBAMP
> software, the aware nodes create overlay links between them and then
> run multicast over the overlay.
> You know that Freifunk was considered for GSoC and one of the projects
> is porting the OBAMP protocol (Overlay Multicast for MANET) from Java
> to C and make an olsr plugin out of it ??
No, I did not know about it.
I'm looking (at the moment) for ways to let the kernel do our multicast
traffic forwarding, instead of doing it in userspace. Maybe I will just
discover "it cannot be done for mesh networks". ^^
> This may be interesting for many developers in the list.
Okay, maybe we should really pull all the people interested in a full (and
efficient) multicast implementation for OLSR together in one place.
> If anybody wants more information can write to me privately. We can
> also have a chance to talk about all this live at the wcw2009 in
> Berlin !
Sounds good :)
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 198 bytes
Desc: This is a digitally signed message part.
More information about the Olsr-dev