Re: [Autoconf] updated draft on aspects of multi-hop wireless communication

["Teco Boot" <[e-mail]>]

Minor comment:
I would compare an OLSR Router with an OSPF Router with an DYMO Router and
with an RIP Router, just to come up with a few protocols. I would not
compare an OSPF Router with an MANET Router, this doesn't make sense, an
OSPF Router could be a MANET Router.
I think MANET is a term for a category of routing protocols, similar to link
state protocols (but an orthogonal category).

Teco.

|-----Oorspronkelijk bericht-----
|Van: [e-mail] [mailto:[e-mail]] Namens
|Thomas Heide Clausen
|Verzonden: dinsdag 24 februari 2009 23:06
|Aan: Rex Buddenberg
|CC: [e-mail]; Emmanuel Baccelli
|Onderwerp: Re: [Autoconf] updated draft on aspects of multi-hop wireless
|communication
|
|
|On Feb 24, 2009, at 20:02 PM, Rex Buddenberg wrote:
|
|> Charlie,
|>
|> Paul has a reality check point and its worthwhile understanding
|> this because missing the point has made both MANET and autoconf
|> harder than they need to be ... IMHO.  It's also made them less
|> relevant to what I think I see in the future.
|> My background is analyzing information systems in DoD and emergency
|> services ... why are they not interoperable (or, in the rare
|> inverse instance, why are they interoperable)?
|>
|> In dealing with some DoD bureaucracy, I'm finding that they don't
|> understand the implications of the differences between LANs and
|> WANs.   And by MANET stating that every end system is also a
|> router, we equally obfuscate the point.
|>
|
|<SNIP>
|
|I think there's a fundamental misunderstanding here. I'll try to
|clarify it, if I can...
|
|In an OSPF-routed network, for all that OSPF cares about, every
|"system is [an OSPF] router".  That doesn't preclude that an OSPF
|router may have interfaces towards other entities, called hosts --
|but as such hosts do not take part in routing (and in the routing
|protocol), they're just not relevant when talking about OSPF.
|
|In a MANET-routed network, all systems that the MANET routing
|protocols care about are.....MANET routers. That doesn't preclude
|that a MANET router may have interfaces towards other entities,
|called hosts -- but as such hosts do not take part in routing (and in
|the routing protocol), they're just not relevant when talking about
|MANETs.
|
|In other words.....it's perfectly fine to hang an Ethernet hub or an
|802.11 access point or whatnot off of a MANET router, assign a prefix
|to that link, hang hosts on that link --  and use the MANET routing
|protocols to exchange that prefix such that these hosts are routable/
|reachable. It's not just perfectly fine, that's what MANET routing
|protocols are build to do ;)
|
|I'd actually make the exact opposite point of the one you're making,
|Rex: every end system is a host -- intermediate systems running MANET
|protocols are routers. Hosts are unaware of if they're hanging off a
|MANET, OSPF, ISIS or other router -- they just see a classic IP link
|and an IP hop and likely a default route. The MANET, OSPF, ISIS or
|other router deals with the "routing stuff", including
|characteristics of links to other routers. Hosts never see that. This
|is as it should be.
|
|Occasionally, a system in a MANET may in the same physical box have a
|logical router and a host  present. This isn't that unusual either
|for non-MANET networks.
|
|So when we talk about systems being routers in MANETs, then it's
|simply because the systems that "we care about" are the routers.
|Hosts are hanging off (some of) these routers just fine, over classic
|IP links -- over which the usual slew of protocols works just fine
|(fortunately -- so we do not have to care about that ;) ).
|
|What MANETs are concerned with are MANET routers and their
|interconnect to other MANET routers. Interconnect from MANET routers
|to hosts is no different from interconnect from a host to, say, an
|OSPF router.
|
|Does this help?
|
|Thomas
|
|_______________________________________________
|Autoconf mailing list
|[e-mail]
|https://www.ietf.org/mailman/listinfo/autoconf