[quagga-users 11420] Re: ripv1 classless routing doubt??

Nick Hilliard nick at inex.ie
Mon Mar 15 13:07:14 GMT 2010


On 15/03/2010 11:58, Krishna Chaitanya wrote:
> So there is no choice but to go with RIP, basically what i am trying to
> do is to validate the devices 1 and 2 in terms of RIP Functionality and
> its RFC Compliance.

--
'On two occasions I have been asked, "Pray, Mr. Babbage, if you put into
the machine wrong figures, will the right answers come out?" I am not able
rightly to apprehend the kind of confusion of ideas that could provoke such
a question.' -- Charles Babbage.
--

Krishna, I don't mean to sound rude here, but unless this is for an
academic assignment, you're on a fool's errand.

You're aware that RIPv1 does not transmit netmasks.  This means that all
RIPv1 updates are assumed to be classful.  This deployment strategy
generally stopped working in 1994, when CIDR routing was introduced.

RIPv1 was formally declared historical in 1996 (rfc 1923).  That's 14 years
ago.

In your case, you're asking what a RIPv1 system will do with
169.254.140.0/24.  This prefix might be transmitted as 169.254.140.0, but
more likely it will be transmitted as 169.254.0.0.  The recipient of this
update also has the freedom to interpret the update any way they want.  So,
this is non-deterministic.  The RFC has nothing to say about this, because
network masks were not relevant in 1988 when the RIPv2 RFC was written.

You're starting out with a broken design.  Whatever you are planning to do
here is deeply misguided and will cause you and/or other people serious
problems because it will not generally work in today's subnetted world.

[In particular, if this is for some form of commercial system, please do
yourself and your customers a huge favour and stop supporting ripv1.]

Nick


More information about the Quagga-users mailing list