Notepad for IETF 121

[BasilGuo]

More problems are raised, but I cannot take the note completely. After watching the session again, I will refine it.

• John Scudder: For information, the BGPSEC design team explicitly considered partial deployability and concluded it was not valuable and was even a negative(because of the illusion of security). But I haven't analyzed the FC-BGP proposal to see if it suffers from the same issues that were considered. I just wanted to make it clear that it wasn't an oversight or accident that BGPSEC doesn't do this.
• Doug Montgomery: https://www.rfc-editor.org/rfc/rfc8374.html#page-253.
• Maria Matejka:
  • consider signatures cost. signing algorithm performance.
  • the complement, attributes in bgp update. looks like one update for each prefix(aggregate).
• Jeffery Haas:
  • bgpsec design issues.fc-bgp attr. 3 asn, bgpsec procedure. code point
  • implementations. open source.
• John Scudder: bgpsec partial deployment. really carefully think through this.
• Alex: aspa, what practice. attr, from aspa?
• Randy Bush: partial deployment. partial path slicing? What you intend to do? partialpath signing. not siging part path.