Re: [manet] draft-ietf-manet-nhdp-10.txt - WGLC - Oct 5th 2009

["Justin Dean" <jdean at itd.nrl.navy.mil>]

Metrics are not currently part of NHDP the way you might be thinking.  There
are up threshold/metric concepts along with state machines which dictate
behavior with regard to these and bringing links up and down but there is no
prebuilt way for calculating these link metrics.  I believe that OLSRv2 will
have some method of doing this (easily done on a two hop basis with
packetbb/NHDP methods) 

Anycast and multicast are currently not supported directly, the issue NHDP
has with this situation is illustrated in my previous example.  If an
anycast/multicast address is included in a hello message with a LOCAL_IF
type TLV (with value either THIS_IF or OTHER_IF) then that address will be
used upon reception to identify the entry in the receiving routers
N_neighbor_addr_list.  This will cause the routers which are transmitting
these addresses to write over each other's "Neighbor Tuple" entry in
receiving routers N_neighbor_addr_list.

A simple fix to this would be with the addition of a different TLV type (say
OTHER_IP), or with a new LOCAL_IF value (say OTHER_IP).  The first case is
cleaner as having a value of OTHER_IP as an entry of the LOCAL_IF list.
Furthermore if a different type was used it could be done as an add on to
NHDP and provide even greater functionality with differing values (say
anycast vs multicast destination vs source).  I am open to thoughts
preferences on this.

Justin

> -----Original Message-----
> From: manet-bounces at ietf.org [mailto:manet-bounces at ietf.org] On Behalf
> Of Teco Boot
> Sent: Thursday, September 17, 2009 2:05 PM
> To: 'Henning Rogge'; manet at ietf.org
> Subject: Re: [manet] draft-ietf-manet-nhdp-10.txt - WGLC - Oct 5th 2009
> 
> Now metrics join the game. Currently not part of NHDP, while this
> is strongly related to neighborhood. Metrics would be used not only
> by OLSR (e.g. distance in DYMO).
> 
> Another thought: is it up to NHDP to keep track metrics for the link
> to the 2-hop neighbor? If not, I think there is no problem.
> 
> And a remark on the address uniqueness: is anycast permitted? And what
> if a lot of duplicates are in the network, all on different media?
> I was present at an event with some ad hoc networking. I saw a lot of
> 192.168.0.0/24 networks. Luckily, nobody was shot.
> 
> Teco.
> 
> |-----Oorspronkelijk bericht-----
> |Van: manet-bounces at ietf.org [mailto:manet-bounces at ietf.org] Namens
> |Henning Rogge
> |Verzonden: donderdag 17 september 2009 18:01
> |Aan: manet at ietf.org
> |Onderwerp: Re: [manet] draft-ietf-manet-nhdp-10.txt - WGLC - Oct 5th
> |2009
> |
> |Am Donnerstag 17 September 2009 17:34:09 schrieb Justin Dean:
> |> An alternate proposal:
> |> NHDP interfaces of a router which running on the same medium must be
> |> uniquely addressed.
> |> NHDP interfaces of a router running on differing
> |> mediums may be addressed the same (their "identifying," THIS_IF,
> |address
> |> sets MUST match completely.)
> |
> |I'm not sure if this will be enough with link quality together (or
> other
> |link
> |specific information in the hellos).
> |
> |Assume a chain of three nodes, A=B-C. There are two different
> sepparate
> |channels between A and B, but only one between B and C (on a third
> |channel).
> |
> |If C receives a hello from B, it cannot tell the link metrics values
> of
> |the
> |two connections between A and B appart, they have the same IP.
> |
> |Maybe I am missing something here ?
> |
> |Henning
> 
> _______________________________________________
> manet mailing list
> manet at ietf.org
> https://www.ietf.org/mailman/listinfo/manet