-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>>>> "Philip" == Philip Levis <[email protected]> writes:
    Philip> I feel like we're running in circles, in part due to
    Philip> different expectations of how RPL will be used. 

    Philip> It's clear that in proprietary or vertically integrated
    Philip> networks running RPL, it's possible to state how
    Philip> controllers/other nodes set the flow label in order to play
    Philip> nice with RPL. E.g., if my home automation network receives
    Philip> commands from a special device (which itself might have a
    Philip> web/SOAP/etc. interface), then that special device can set
    Philip> the flow label. This isn't the problem case. 

    Philip> The problem case for the use of the flow label is when
    Philip> arbitrary Internet hosts (e.g., my laptop) want to
    Philip> communicate with devices running RPL. E.g., I have a home
    Philip> RPL network and want to directly talk to a RPL device from
    Philip> my laptop. Requiring the packet source to be aware of flow
    Philip> labels and how to set them breaks the whole idea of Internet
    Philip> connectivity: suddenly an end host needs to be aware of what
    Philip> routing protocol is used to reach the other host, as well as
    Philip> configuration parameters of that host. I need to be able to
    Philip> SSH to the RPL-connected device without requiring a new ssh
    Philip> program that uses ioctls to set flow labels. We can't say
    Philip> that only OSes which set the flow label to 0 (or some other
    Philip> value) can be used. Nor can we expect to change how OSes set
    Philip> the flow label. 

I guess I don't see the problem to be as big a deal as you suggest.
I'm happy if flow label 0 gets some default RPLinstanceID.  

It would be convenient if the rules were relaxed such that on ingress,
the RPL edge router could *set* the FL if it were zero, and it appears
that this matches up well with what the load balancing people want to
do.

Today, if you ssh or HTTP or what-not, without doing anything special,
you will get a flow label of zero, which with the above rule, means you
get connectivity.  Only if some intermediate node sets the FL=0 would it
be a problem today, and that's illegal so far.

BTW: the flowlabel is not in an ioctl, it's right in the sockaddr_in6.
Too bad that inet_ntop(3)/RFC2373/4291 was not specified to include the flow
label, as then it would be just a matter of user education...  Maybe
there is some way to use the %ifname notation used for ifindex mapping.
(okay, it falls apart once you want to use DNS)

It seems to me that within RPL, if we use RPLinstanceID, we need to
decide what to do when we get a flow label from an untrusted source that
is non-zero.  Inserting an end-host option, with the unstrusted FL
seems like the right thing to do to me.  This facility might be
generally useful to entire Internet.

- -- 
]       He who is tired of Weird Al is tired of life!           |  firewalls  [
]   Michael Richardson, Sandelman Software Works, Ottawa, ON    |net architect[
] [email protected] http://www.sandelman.ottawa.on.ca/ |device driver[
   Kyoto Plus: watch the video <http://www.youtube.com/watch?v=kzx1ycLXQSE>
                       then sign the petition. 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Finger me for keys

iQEVAwUBTGGx4oCLcPvd0N1lAQIuhgf/Z6e/vJKj9K+6+N3djKdLx7nut2iJ+K1X
JaKs/TsKJdNJ59520x9m5pcYo+r9oXrGVG2mohIHu0WktY/XkJMdBSncgkiKvSmh
VEZlFAzw+eJYZdiNlWvIFxD13REOyc4S+jJhx6ECONFSq2csHdbt8q7O1jIsDuar
hEhLS7v0a61T7rbM7aP95q+aYBr2ofWX5g596I/lYgSwZ7+fhuySm7DhckBWf/yC
pPdEw4xAMA4/jnHZzp3iTKcFhbGQWvdBvsUKPM/EJLQWwExzFS3WmEP3ltDyXkiZ
zo5oEj7Xdir2mN18cvyysa/8izgQ8sD3TSYUu+feNHO2hRdbNRoX3w==
=kE6H
-----END PGP SIGNATURE-----
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to