Hi Uma,
On 8/26/15 07:59 , Uma Chunduri wrote:
But I would note - maintaining unique indices across all nodes
(starting from the starting point of the SRGB) for each topology is
way harder than maintaining one set of unique index per node. With 10
topologies in the below example, with 1000 nodes per topology, in per
index case operator ought to maintain 10 * 1000 = 10000 unique
objects. But in the other case of per topology SRGB operator has to
only maintain
1000 (unique labels/SID Index) + 10 (SRGB offsets) = 1010 unique objects in
total. I am not sure how maintaining
*more* unique objects would be easier.
[Les:] Regardless of approach, you have a unique label for a given
prefix/topology pair.
So you have the same number of "objects" in both cases. This is very clear if
you look at
what is installed in the forwarding plane. Do not make the mistake of confusing
CLI
syntax w the number of labels being used.
[Uma]: Les, No- I was not making any mistake of confusing with CLI
syntax. Also, I was not at
all taking about uniqueness from forwarding plane PoV. In either
approach those
of course would be unique from forwarding plane PoV.
The point is maintaining unique SID values per topology and per
algorithm would be from
operational standpoint is excessively difficult (as we can see unique
objects to be
provisioned would be multiplied) and easily avoidable.
right, but to do that you do not need per topology SRGB.
thanks,
Peter
--
Uma C.
_______________________________________________
spring mailing list
spring@ietf.org
https://www.ietf.org/mailman/listinfo/spring
