On Wed, Feb 24, 2010 at 12:04:25AM +0100, Ole Troan wrote:
> I took the SHOULD from RFC3315, which has that for the ORO option. I
> couldn't find any other DHCP option RFCs with SHOULD or MUST for the
> PRL option.

I think many option RFC's do not mention the PRL explicitly (how the
client gets the option is a basic DHCP protocol feature in that
sense), but in this case I'm a little uncomfortable because the
language of the MUST about there being one option is so specific.

I think in RFC 3315's case the SHOULD is because the client can elect
not to send the ORO at all.

I wonder if the reason that there be at most one option is because you
may be aware of 'option fragmenting'?  In DHCPv4, multiple options of
the same option code are concatenated together in order to support
options whose data is longer than 255 octets (and it is also handy in
letting smaller options be split along the options/file/sname fields
if out of space and using option overloading, but these details are
why I don't like busying a new option specification with this
discussion; it gets confusing fast).


It may be simpler to say;

  Because the OPTION_6RD contains one IPv4MaskLen/6rdPrefixLen/6rdPrefix
  block, and because DHCP cannot convey more than one instance of an
  option, OPTION_6RD is limited to provision at most a single 6rd
  domain.  Provisioning of a CE router connected to multiple 6rd domains
  is outside the scope of this protocol specification.

And avoid the normative language entirely (leave all those mechanics
entirely up to RFC 2131 conformance)?


Anyway, it is just an idea.  I'm quite happy with there just being a
reminder about the PRL with the existing text.

-- 
David W. Hankins        BIND 10 needs more DHCP voices.
Software Engineer               There just aren't enough in our heads.
Internet Systems Consortium, Inc.               http://bind10.isc.org/

Attachment: pgptHO96rtlH6.pgp
Description: PGP signature

_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires

Reply via email to