Re: mboned: RE: suggestion for simplifying IGMPV3/MLDV2 to discard EXCLUDEmode
Marshall Eubanks <[email protected]>
| Newsgroups | gmane.ietf.magma,gmane.ietf.mboned |
|---|---|
| Message-ID | <[email protected]> |
Hello; On Jun 29, 2006, at 1:26 PM, mark fine wrote: >>> I thought the desired outcome of this discussion was to >>> eliminate the need for EXCLUDE filter state on the router. >>> Am I mistaken here? >> >> That was my understanding. > > draft-holbrook-idmr-igmpv3-ssm completely eliminates the need for > EXCLUDE > filter state (it does not allow IS_EX or TO_EX messages) at the > expense of > not providing an ASM service. i thought the desired outcome was to > remove > excluded source lists while still allowing (*,G) joins, as well as > included source lists. eliminating excluded source lists > significantly > reduces / neuters the EXCLUDE router filter mode to a concept that > only > supports (*,G) state (as well as tracking requested included source > lists): whether it's called EXCLUDE filter mode or renamed to > something > like ASM mode, it's effectively a group record with an active timer > and > represents a (*,G) join. this accomplishes the goal of providing > both SSM > and ASM services within the interoperable context of the full > IGMPv3/MLDv2 > protocols, while dropping the ability to exclude source lists. > > that's my understanding :) Yes. And one question is : Does anyone know of an application / service that actually requires tracking exclude source lists ? I thought once (around 2000-2001) that this could be used to squelch DOS attacks on ASM groups, by excluding the rouge actor, but I think that both the usage and the technology have evolved in different directions so I don't see removing that possibility as a real issue. > > mark > > Regards Marshall