[rbridge] Queries on RFC 6325

[Donald Eastlake]

Hi Siva,

On Tue, Nov 15, 2011 at 2:05 AM, Sivakumar V <sivakmv at gmail.com> wrote:
> Hi Authors,
>
> 1. I was going through RFC 6325 section 4.5.4 Tree Distribution optimization
> section.  RFC tells about attachment state of vlans and routers, which was
> not clear to me ,  as I have no idea of attachment state of vlans and
> routers.So would you please throw a light on attachment state.

An RBridge is considered to be "attached" to a VLAN if it is Appointed
Forwarder for that VLAN for any port of that RBridge.

Layer 3 routers are end stations. I think that Section talks only
about multicast router attachment, which is what it sounds like. One
of the RBridges ports is connect to a link that also has a Layer 3
(actually IPv4 or IPv6 or both) multicast router attached.

> 2. In section of 4.5.2 Multi-Destination Frame checks, RFC tells about
> potential ingress trees,  in which I am not clear about the implicit
> part:"The j potential ingress trees for RB2 are the ones with nicknames that
> RB2 has explicitly specified "specified ingress tree nicknames" (and that
> are included in the k campus-wide trees selected by highest priority RBridge
> RB1), with the remainder (upto the maximum of {j,k}) being the highest
> priority of the k campus-wide trees."
>
>  suppose RB2,not the highest priority rbridge, has specified j=7 and
>
> D3,D4 are explicitly  "specified ingress tree nicknames" .RB1, highest
> priority RBridge,has specified K=5 and D1,D2,D3,D4,D5 as campus wide trees
> to be calculated by all the rbridges in the campus.
>
>  RB3 has to identify potential ingress trees of RB2,ingress trees are trees
> which are explicitly specified by RB2 i.e. is D3,D4 and remainder are
> implicit ingress trees upto 7 (as 7 is maximum
>
> {j=7,k=5}) are taken from k-campus wide trees which are  D1,D2,D5.But the
> concern here is only 5 trees have been selcted D1,D2,D3,D4,D5 instead of 7.
> From where remaining 2 trees are selected ? can you please clarify on this.

Sorry if the wording is a little unclear but there are lots of
exceptions and corner cases here...

There is a total ordering of all nicknames as tree roots. This is
ordered by priority to be a tree root except that the RBridge that is
the highest priority to be a tree root can alter the order by
specifying some nicknames to be placed first in that ordering. When
some other RBridge specifies the trees it can use, it takes the top j
items for this ordered list except that it, also, may want to
prioritize the use of certain trees. So the other RBridge can list
trees that it wants to use. But if it doesn't list enough, the rest
are taken from remaining nicknames on the total ordering list.

Anyway, I believe that in the case you describe above, the 6th and 7th
trees would be whatever two trees are highest in priority in the
campus wide list after removing the D1 through D5. If, for example,
there are only five nicknames in the campus, then it doesn't matter
that j=7, only 5 trees would be possible so only 5 trees would be
used.

Thanks,
Donald
=============================
 Donald E. Eastlake 3rd   +1-508-333-2270 (cell)
 155 Beaver Street, Milford, MA 01757 USA
 d3e3e3 at gmail.com

> Regards,
>
> Siva