Re: [Roll] [roll] #128: Trickle multicast could be considered in other applications?

["Dijk, Esko" <[e-mail]>]

Don,

I agree with you on the role of the multicast scope. MPL is defined well enough for this aspect. However I was referring to a different situation than you had in mind, most likely.

The situation I have in mind is a LLN mesh network where 'hosts' (IP hosts) can reside at any position in the network. For example, a mesh network where 50% of nodes are MPL Forwarders (together constituting a single MPL domain) and 50% of nodes are hosts that don't know anything about MPL. Suppose one of these hosts, host A, sends out an IP multicast packet and three nearby MPL Forwarder nodes receive this multicast. Then, MPL-04 does not describe whether these MPL forwarders, or which of these, are required to "inject" the multicast packet into the MPL domain. MPL-04 does say that an MPL Forwarder may "inject" such a packet, acting as an MPL Seed.
Similarly there's also no specification how a host B that is 5 hops away from host A, could ever receive this multicast packet. (This won't happen automatically since host B does not listen to the ALL_MPL_FORWARDERS address - it doesn't know about MPL. There's no requirement that every MPL Forwarder "decapsulates" the received packets and resends them unencapsulated for the benefit of hosts.)

I think Section 3 could just mention that additional mechanisms for the above are out of scope.

regards,
Esko


________________________________
The information contained in this message may be confidential and legally protected under applicable law. The message is intended solely for the addressee(s). If you are not the intended recipient, you are hereby notified that any use, forwarding, dissemination, or reproduction of this message is strictly prohibited and may be unlawful. If you are not the intended recipient, please contact the sender by return e-mail and destroy all copies of the original message.