From Yoshifumi Nishida on the [IETF Multipath ML](https://mailarchive.ietf.org/arch/msg/multipathtcp/qXcTQIAaH3ijjh7dP0xGkyQt64A/) > I think a host that sends NEW_KEY option needs to give up resending it at some point. > I think the draft needs to address this point.