Re: [riot-devel] How to Choose Network Interface to Send Packets Over

2015-11-05 Thread Mark Solters
[Sorry if this message ends up in the wrong place or as a separate thread
-- I was stuck in digest mode and it was not clear to me how to reply to a
single message.]

Oleg, thank you very much for the description of and reference to the
packet routing protocol -- I did not realize that RIOT followed the
contents of the RFC document so precisely.

I now know that the process of deciding what network interface to route a
packet to comes down to (at first) finding an on-link match between the
destination IP address and the available interfaces (if one exists).  This
is done by examining the prefixes for a match.

In my case, I have two interfaces, ethernet and 802.15.4, with
auto-assigned local IPV6 addresses:

[Iface 6, 802.15.4] = fe80::5846:346f:c2da:ad12/64
[Iface 7, ethernet] = fe80::204:25ff:fe19:1a3e/64

So I think my problem is as follows.  I am transmitting packets to a
destination IP that reads:

fe80::3a2c:4aff:fe6e:19f3

And therefore the fe80:: could match the 64-bit prefix of *either* of those
interfaces.  I think that the request is then being routed to the first
match, which is 802.15.4, which is incorrect, because that IP address is
actually an ethernet neighbor, not a radio one.  When I disable Iface 6,
the packets DO reach their destination on Iface 7.

Maybe I'm wrong; but it would seem that my problem is, I need to change the
prefix of my 802.15.4 interfaces to something other than fe80:: so they do
not conflict with the ethernet interface's prefix, which is beyond my
control.

Am I on the right track here?  I do not as of yet fully comprehend the
addition of the global affe::2 address in the gnrc_border_router example;
and additionally, I am uncertain if my reasoning here is still flawed due
to a deeper failure to internalize the core concept behind RIOT's
implementation of the packet routing algorithm.

Many thanks,

Mark
___
devel mailing list
devel@riot-os.org
https://lists.riot-os.org/mailman/listinfo/devel


Re: [riot-devel] How to Choose Network Interface to Send Packets Over

2015-11-03 Thread Oleg Hahm
Hey Mark!

> This leads me to a slightly deeper question -- how does RIOT decide what
> physical interface to attempt to communicate over when I use network send
> calls such as conn_udp_sendto?

Selecting the proper interface for sending, is task of the network layer and
is implemented according to the procedures explained in the according RFCs
[1][2] in GNRC. This means that a node will first check if the destination
prefix is "on-link" (if applicable) and use the according interface. If not,
it will look up the destination in the FIB (Forwarding Information Base) if
available (e.g., if a routing protocol is available or someone else filled the
FIB manually) for the next hop's IP address. You can check the content of the
FIB using the `fibroute` shell command. Finally, the neighbor cache (check
with `ncache` shell command) is checked or a neighbor solicitation may be
performed.

Hence, if your IPv6 prefixes are configured correctly (e.g. as described in
the gnrc_border_router example's README), the appropriate interface should be
chosen.

> As an associated question -- is it the intention of the gnrc_border_router
> example to automatically bridge interfaces (such as ethernet & 802.15.4)?
> As far as I can tell it really only initializes interfaces but does not
> provide any actual routing between them.

The border router is (as the name suggests ;-)) a router, not a bridge.
Forwarding packets via the correct interface is done according to the rules
stated above.

HTH,
Oleg

[1] https://tools.ietf.org/html/rfc4861#section-5.2
[2] https://tools.ietf.org/html/rfc6775#section-5.6
-- 
I'd make a joke about UDP, but I don't know if anyone's actually listening...


signature.asc
Description: PGP signature
___
devel mailing list
devel@riot-os.org
https://lists.riot-os.org/mailman/listinfo/devel