Re: [Idr] Progressing draft-bhatia-bgp-multiple-next-hops-01.txt

[Paul Jakma <paul at clubi.ie>]

On Thu, 24 Aug 2006, Enke Chen wrote:

> The update sequence described by Manav is the following:
>
> UPDATE: feasible nlri X, next_hop N1, <other attributes>
> UPDATE: feasible nlri X, multiple_hop N2, <other attributes>
> UPDATE: unfeasible nlri X, multiple_hop N1
> UPDATE: unfeasible nlri X, multiple_hop N2
> UPDATE: feasible nlri X, multiple_hop N11, <other attributes>
> UPDATE: feasible nlri X, multiple_hop N12, <other attributes>

> You seem to be talking about a different approach. Could you detail the 
> sequence of updates?

Err, you may have meant "detail the sequence for the other approach" 
rather than "detail the sequence above". I'm not sure what other 
approach you mean, unless you refer to the opportunistic bunching 
that is possible when, for a specific NLRI, the paths differ only by 
nexthop.

E.g. if for all the above <other attributes> are equal, then it 
is possible to 'bunch' some of the UPDATEs together:

- the first two UPDATEs differ only by nexthop, so they can be
  opportunistically bunched together into:

UPDATE: feasible nlri X,
        next_hop N1,
	multiple_hop N2
	<other attributes>

- The second two explicit withdrawal UPDATEs can be bunched together:

UPDATE: unfeasible nlri X,
	multiple_hop N1, N2

- The last two UPDATEs can be bunched together:

UPDATE: feasible nlri X
	multiple_hop N11, N12
	<other attributes>

Whether it's possible obviously depends on timing, as it does today 
for UPDATE bunching.

regards,
--
Paul Jakma	paul at clubi.ie	paul at jakma.org	Key ID: 64A2FF6A
Fortune:
petribar:
	Any sun-bleached prehistoric candy that has been sitting in
	the window of a vending machine too long.
		-- Rich Hall, "Sniglets"

_______________________________________________
Idr mailing list
Idr at ietf.org
https://www1.ietf.org/mailman/listinfo/idr