> On 5 Apr 2019, at 20:15, Michael Richardson <[email protected]> wrote: > > If we could delay the switch to the new schedule until *after* > COJP_REKEYING_GUARD_TIME has elapsed, then that might work. > Alternatively, if we could number the rekeys then a EB could signal the > switch to the new schedule some time after the keys are changed.
Not sure about that the exact switch after COJP_REKEYING_GUARD_TIME would work, different nodes receive the new keys at different instants so it would be hard to sync them. I like better the trigger with an EB, it originates at 6LBR that is trusted and the frame can be authenticated by the nodes that have installed the new keys. Mališa _______________________________________________ 6tisch mailing list [email protected] https://www.ietf.org/mailman/listinfo/6tisch
