Minutes for ROLL at IETF-interim-2010-roll-1
minutes-interim-2010-roll-1-00

Minutes from interim ROLLs Meeting
Monday 20, December, 2010
Time: 8:00am PST = Noon ET = 6pm CET

Thanks to Dan King for Taking the Notes

20 or so in attendance via WebEx

1. ROLL WG update
Current drafts
- draft-ietf-roll-routing-metrics is in Last Call.
- Working to clear IESG discussed on draft-ietf-roll-rpl.

2. Recharter Discussion
- JP> Plan is to continue the discussion that we started at the last IETF util
end of January and come back to the WG with a summary by end of January 2010
Management (COAP SNMP, et al) model need to be discussed - It was stated that
access points are rarely managed via SNMP and instead primarily managed via web
interfaces. - Co-chairs will also look for assistance from AD's to help review
the management requirements and identify management experts as needed.
Lightweight crypto algorithms - Some IPR issues exist with existing solutions
so should the working group await a legal resolution, or attempt to create
alternative solution? The co-chairs are requesting people to respond with their
comments on the WG list. Label Switching - It was agreed that label switching,
source routing and flow switching technologies may be relevant for the ROLL WG.
Further discussion and identification of the requirements and debate of
potential solutions is needed. This may be kicked off with a requirements and
applicability draft. - Generally the IETF WGs responsible for label switching
and associated technologies are CCAMP, MPLS and PCE. - Anyone interested in
this area may find RFC4728 (The Dynamic Source Routing Protocol (DSR) for
Mobile Ad Hoc Networks for IPv4) of interest.

RPL on “Not So Lossy Links”
- An item that was mentioned previously and requires further discussion. If you
have an opinion, please respond to the WG list.

3. RPL Enhancements (for new charter)
Presenter Jorjeta Jetcheva
Multicast Enhancements
- draft-hui-6man-trickle-mcast-00 was proposed previously, although not ROLL
specific. - Due to the nature of sensor/seed networks, issues exist on which
devices need to be listening and/or scheduled for wake-up. - The mechanisms for
multicast do not necessarily need to be link-layer based. - Co-chairs would
like to see discussion on multicast applications and requirements via the WG
mailing list. Mechanisms for Controlling DODAG attachment - Requirements exist
for discouraging or rejecting a node from attaching to a DODAG. - Discussion is
also needed for routing admission control and load sharing. JP (co-chair hat
off) agreed that this is a useful feature to support - Requirements and
experiences are requested from deployed networks. Neighbor Adjacency
Maintenance - Need to be cautious as maintaining routing adjacencies is quite
expensive. Originally the requirement was not to support this in RPL. Actual
requirements need further discussion. - The WG was asked to consider how
maintenance mechanisms might be implemented and the benefits. Any mechanism may
or may not require maintaining the state of a neighbor, and or switching to
another (due to down state) neighbor. - It was highlighted that it may be worth
including how we discovery link metrics between neighbors. - The WG should also
be careful not to duplicate mechanisms for closely related activities.