On Wed, Feb 25, 2015 at 2:45 PM, Joakim Gebart <[email protected]> wrote: > Hello list, > I have been trying to get Contiki nodes to connect to a RIOT RPL > network and vice versa, but so far without success. By sniffing the > wireless traffic I have found that Contiki seem to be sending its RPL > messages to the multicast group FF02::1A which in [1] is called "All > RPL nodes". RIOT on the other hand is sending its RPL messages to > FF02::1 and FF02::2 which are the "All nodes" and "All routers" groups > which seem to be ignored by Contiki. RIOT seem to be unaware of the > FF02::1A group on the other hand. > > My question is: What is the correct behaviour for RPL messages? > > Is there a defined multicast address to use (according to some RFC > document or whatever)?
Actually, to respond to my own question: I found this in RFC6550: Most RPL control messages have the scope of a link. The only exception is for the DAO / DAO-ACK messages in Non-Storing mode, which are exchanged using a unicast address over multiple hops and thus uses global or unique-local addresses for both the source and destination addresses. For all other RPL control messages, the source address is a link-local address, and the destination address is either the all-RPL-nodes multicast address or a link-local unicast address of the destination. The all-RPL-nodes multicast address is a new address with a value of ff02::1a. RIOT's behaviour in this regard therefore seems wrong. http://tools.ietf.org/html/rfc6550#section-6 > > [1]: > http://www.iana.org/assignments/ipv6-multicast-addresses/ipv6-multicast-addresses.xhtml#link-local > > Best regards, > Joakim Gebart > Managing Director > Eistec AB > > Aurorum 1C > 977 75 Luleå > Tel: +46(0)730-65 13 83 > [email protected] > www.eistec.se _______________________________________________ users mailing list [email protected] http://lists.riot-os.org/mailman/listinfo/users
