Hi Ali This draft seems to be a further optimization of L3 optimized IRB draft below. Please correct me on my understanding.
https://datatracker.ietf.org/doc/html/draft-sajassi-bess-evpn-l3-optimized-irb-04 So the above draft optimizes L2fib to only contain the locally connected to reduce resources for PE and CE ARP/ND cache. This draft provides a further EVPN optimization with a new S bit slice bit added to default gateway extended community to the RT5 summary prefix advertised from aggregation node to lightweight constrained edge PE leaf or tor nodes. In cases where EVPN L2 VNI is being stretched between PODs using multisite or RFC 9014 DCI overlay and in that case can the anchor PEs exist in a different POD that is being extended. Also can this also be applied to BGP Only DC? So with a combination of both drafts being layered on top of each other for an operator implementation the the L3 optimized IRB with optimized flag limits the PE to ND/ARP L2FIB to local routes and CE to the same with addition of DAG Anycast MAC. With this subnet slice draft layered on top the lightweight leaf or tor only has the local L3RIB L3FIB for RT2 plus RT5 summary route from anchor PE with S bit set in the default gateway extended community. Kind Regards Gyan
_______________________________________________ BESS mailing list -- [email protected] To unsubscribe send an email to [email protected]
