If by router up/down you mean only the case of the overload bit changing, then
yes, if section two states those listed cases are the applicable events I
I would understand the value and the behavioral tradeoff.
At 05:30 PM 11/21/2005, Olivier Bonaventure wrote:
> Remember that "protected by a fast reroute protection technique" may be
> on a prefix by prefix basis (protected for some, but not for others) and
> may be hard to determine.
Agreed, this depends on the type of protection technique. A link should
be marked as protected only if all the destinations are protected.
> If this were restricted to
> o link up
> o metric change
> then I would not be as uncomfortable. Those are indeed generally cases
> where care is more important than speed.
> Joel M. Halpern
> PS: Orderly shutdown would then be a manual metric change followed later
> (a few seconds later would suffice) by abrupt removal.
Would you agree if section 2 of the draft was updated and link down
events were removed and only the following events were supported :
- metric change
- link up
- router down and router up (overload bit transition in ISIS)
This would mean that if a router wants to either supports graceful
manual link shutdown or graceful shutdown of fully protected link, it
should first advertise the considered link with a very large metric and
later remove it from its link state packet.
Rtgwg mailing list