Authors,

As part of shepherd review, please find some comments.

Editorial

  *   “When the the source” → “When the source”.
  *   “Singe-Active” → “Single-Active”.
  *   “on-touched” → “untouched”.
  *
“ethernet segment” → “Ethernet Segment”.
  *   “Section 6 and 7” → “Sections 6 and 7”.

Minor :

  *
Section 10 SEMG behavior uses several lowercase “should/must” statements where 
protocol behavior seems required, especially discarding SMET/IGMP from non-DF 
SEMG and non-SEMG PEs. These should be reviewed for RFC 2119 capitalization.
  *
Missing reference for IEEE802.1Q and unused RFC7348 reference.

Major to consider :

  *
Capability signaling should probably be normative : Section 6 says seamless 
interop capable PEs SHOULD attach the EVPN Multicast Flags EC with “EVPN/MVPN 
Seamless Interop Supported” set. For mixed networks, this seems important for 
safe behavior. Consider whether this should be MUST.
  *
All-active intra-ES tunnel behavior needs tighter normative language :Section 
6.1 says traffic received from inter-subnet tunnel “should not” be sent to the 
Intra-ES subnet tunnel. This is a loop/duplication prevention rule and should 
likely be MUST NOT.
  *
Security Considerations are too thin : Section 15 only inherits security 
considerations from RFC7432/RFC6513/RFC6514/RFC9251. Since this draft 
introduces new interworking behavior, VRI use with EVPN routes, SEMG behavior, 
and possible traffic attraction, it should mention risks from 
incorrect/malicious route import, forged MVPN/EVPN routes, state exhaustion, 
and misconfigured gateways.


Mankamana
_______________________________________________
BESS mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to