Hi Bill,
|> B1d. Semi-hierarchical numeric LOCs are dynamically assigned. Local
|> reconnection during link state changes is accomplished with rerouting
|and /
|> or renumbering.
|>
|> I think this variant is missing, and it is an important one (e.g. ad
|hoc
|> networks).
|
|Is it fully dynamic (routers and subnets are dynamically assigned in
|some kind of hierarchy on up to the core) or is it a combination of
|dynamic and administrative (LAN addresses are dynamically assigned
|while router addresses and subnets are fixed)?
|
|If the former, what makes it different from B1a? Or to put it another
|way, could B1a be reasonably rewritten to include it?

B1a mentions "supernet" upstream and "subnet" downstream. This is definitely
not the case in B1d. In B1d, the IGP can route for any assigned prefix in
the edge network. Failures in the path to the BR that corresponds with the
SA is solved by rerouting.
Rerouting could be used if one of a set links to an ISP fails, or one of a
set middleboxes fails. Renumbering is needed when no path to an ISP is
available anymore. Smooth renumbering (using metrics) is supported to take
access to ISP out of service, or path becomes less optimal, e.g. when
wireless links are used.
The model has 2 layers: the path is inner-edge or edge-BR-core-BR-edge.
Edge-core traffic is allowed (e.g. ICMP).


|If the latter, what makes it different from B1c?

I am not sure what is meant with "administratively assigned". In the B1d
strategy, nothing in the edge needs manual configuration. The BR (part of
edge network, attached to ISP) gets prefixes from ISP with some protocol
(e.g. DHCP-PD or AAA). All routers autoconfigure addresses (e.g. for
management) and get prefixes. Routers come and go. This is the ad hoc type
of operation, to be addressed in Autoconf.


On rewording B1a: Maybe some proposals suggest hierarchical prefix
delegation. This is very different from the PD with IGP approach. If B1a is
reworded for PD with IGP, one could say a variant of a strategy is missing
(again...).


Teco.

_______________________________________________
rrg mailing list
[email protected]
https://www.irtf.org/mailman/listinfo/rrg

Reply via email to