Re: igmpv3/mldv2 equivalent messages

mark fine <[email protected]>
Newsgroups gmane.ietf.magma
Message-ID <Pine.GSO.4.64.0610021137090.6307@modena>
TO_EX({}) and IS_EX({}) messages are equivalent and produce equivalent
operations in the router state tables:

    INCLUDE (A)    IS_EX ({})   EXCLUDE ({}, {})    Delete (A)
                                                    Group Timer=GMI

    EXCLUDE (X,Y)  IS_EX ({})   EXCLUDE ({}, {})    Delete (X)
                                                    Delete (Y)
                                                    Group Timer=GMI

    INCLUDE (A)    TO_EX ({})   EXCLUDE ({}, {})    Delete (A)
                                                    Group Timer=GMI

    EXCLUDE (X,Y)  TO_EX ({})   EXCLUDE ({}, {})    Delete (X)
                                                    Delete (Y)
                                                    Group Timer=GMI

with that said,

a) IS_EX messages are solicited report records sent in response to
    queries, TO_EX messages are unsolicited reports records sent in
    response to interface state changes

b) IGMPv2 / IGMPv1 / MLDv1 reports are BOTH solicited reports sent in
    response to queries and unsolicited reports sent in response to
    interface state changes

c) IGMPv2 / MLDv1 leaves are unsolicited reports sent in response to
    interface state changes

since IGMPv2 / MLDv1 leaves are used only in unsolicited manners, it
is straightforward to translate them to unsolicited TO_IN({})
messages.  but IGMPv2 / IGMPv1 / MLDv1 reports are used in both
unsolicited and solicited manners, so it is possible to translate them
to either unsolicited TO_EX({}) messages or solicited IS_EX({})
messages.  to reduce unnecessary complexity and to standardize on a
single message mapping, and in light of the protocol equivalence of
the TO_EX({}) and IS_EX({}) messages, i imagine that the more common 
solicited IS_EX({}) messages (query responses should
typically outnumber
interface state changes) were chosen as the translation for IGMPv2 /
IGMPv1 / MLDv1 reports.

that's my understanding...

mark

On Mon, 2 Oct 2006, Hitoshi Asaeda wrote:

> Hi all,
>
> I have a question about multicast router behavior for older version
> compatibility.
>
> Regarding the message translation, 7.3.2 of RFC3376 says that IGMPv3
> router recognizes v1/v2 Report to IS_EX().
>
>       IGMPv2 Message                IGMPv3 Equivalent
>       --------------                -----------------
>
>         Report                        IS_EX( {} )   <--- TO_EX()?
>
>         Leave                         TO_IN( {} )
>
>       IGMP Message                  IGMPv3 Equivalent
>       ------------                  -----------------
>
>         v1 Report                      IS_EX( {} )  <--- TO_EX()?
>
>         v2 Report                      IS_EX( {} )  <--- TO_EX()?
>
> As well, 8.3.2 of RFC3810 says;
>
>   MLDv1 Message                 MLDv2 Equivalent
>   -------------                 ----------------
>
>      Report                        IS_EX( {} )      <--- TO_EX()?
>
>      Done                          TO_IN( {} )
>
> IS_IN and IS_EX are transmitted as the response of query. I think
> these join report messages should be translated as TO_EX(), not
> IS_EX(). It is reasonable that leave should be translated as TO_IN()
> as mentioned in RFC.
>
> Am I misunderstanding?
> --
> Hitoshi Asaeda
>
> _______________________________________________
> magma mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/magma
>
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.