Rare project developers

[mc36 <>]

heyyy!
interesting ideas and not too hard to have afterall...
at least now i see why the rift chairs said that zte maybe have an 
implementation... :)
"our collegues" tells me you know some of the authors? if yes, maybe worth 
writing a mail to them....
hopefully they'll participate in an interop test, i'm very curious about the 
whole topic:
so right now, we don't limit the east-west perfix tie advertisement and i do 
2 full spf on the tie database,
so in our implementation, this arbitrary ring/chain of leafs is already a 
there: that's why i badly want the
rift-sr extensions to mature, because with this in mind, one can use rift 
instead of ospf or isis for example,
but with the additional possibility of the inter-level aggregation... you can 
think about this whole thing as if
you have isis/ospf level1/area123 and level2/area0, but if the need arise, 
you can configure an even upper level
above current level2/area0, lets say level3/area-1 to aggregate two or more 
independent level2/area0s... and you
have 24 levels of this hierarchy in hand... so it can be a huuuuge network 
and the fib still can remain reasonably
small as you go down toward the leaves... for the zte guys, we can combine 
this with the gtp-u termination in asic... :)
br,
cs


On 1/5/23 13:55, Simon Leinen wrote:
> Hi Csaba,
>
> I just stumbled over this I-D from our colleagues at ZTE.  It proposes a
> small extension of the RIFT protocol for some strange topologies.
>
> Looks like the ZTE folks want to use RIFT for IP-based RANs (Radio
> Access Networks)... this looks a bit weird to me, since RIFT was
> designed for certain very regular topologies.  But hey, why not? :-)
>
> Cheers,