[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [RRG] Tunnel fragmentation/reassembly for RRG map-and-encaps architectures
If we assume that ETR addresses (unicast or anycast) are in the BGP
as today, and agreed that there would be much less of them then EIDs
why LISP (presumably you are referring to the data plane idea) has
to convey if ETR is up ?
Because if one of the failure cases I mentioned occurred, you want a
source-site to use *another* locator from the locator-set that is
mapped from the EID-prefix for the destination site.
Am I right that the assumption is that the major reason for this is
that you do not want to run BGP on ETRs ? You would not need to ...
You could even run BFD between PE-CE/ETR to indicate to PE that CE/
ETR is down. PE would then re propagate this fast in BGP.
No, that's not the reason. Eliminating BGP from the CE routers is a
side benefit. But if you want to choose locators based on BGP policy,
the CE routers must run BGP.
I am just trying to see a consistency in reachability propagation
end to end.
Understand, no problem.
BGP may say ETR is up while LISP data plane idea that is went down.
Actually for unicast
BGP does not specifically convey ETR reachability. It conveys a prefix
that includes a bunch of ETR addresses. So if an ETR goes down, the
prefix won't change in the core, so the fact the ETR went down cannot
be propagated far away. This is a good thing IMO. But the ETR failure
only goes to the place where the "more-specific route" travels and
then suppressed by prefix aggregation.
I know I am preaching to the choir on this point.
this may work well but for anycast ETR address I am not that sure
during the failure of one ETR from given anycast group.
Well for an anycast ETR that goes away, most of world thinks nothing
happens. Only the place close to the failure detects "the new path to
the address" where, in turn, is actually a new place and not just a
new path. Again preaching to the choir.
Dino
--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg