[dhcwg] Minutes from Yokohama meeting

DHC WG

These minutes were prepared from notes taken by Stuart Cheshire and
Ted Lemon.

Status report
-------------
Ralph started the meeting with a WG status report.

Ralph announced the publication of "DHCP Auto-configure Option
(Option code 116) Deprecated" <draft-droms-rfc2563-deprecate-00.txt>.
It will go to WG last call after publication of
<draft-zeroconf-ip4-linkloocal-xx.txt> as Proposed Standard.

Matt Osman, <draft-ietf-dhc-packetcable-02.txt>
-----------------------------------------------
Matt Osman presented the "DHCP Option for CableLabs Client
Configuration".  This option will redefine the option 177 currently in
use by CableLabs.

Q&A:
Narten: Why not use vendor-specific options?
Droms: Device can only have one vendor identifier; can't specify both
vendor and "CableLabs-compatible" as vendor type.
Droms: Two issues from chair's point of view: CableLabs wants to work
with IETF to generate Internet Standard for function now deployed in
option 177 (note that 177 is from the site-defined option code space);
this draft proposes a somewhat different mechanism for sub-option code
assignment.
Narten: Does this option fundamentally change the way DHCP works?  WG
members should consider that issue when reading the draft.
Woundy: CableLabs is not looking for IETF/DHCWG to rubberstamp this
draft.  CableLabs wants to go through IETF process and standardize the
use of this option.

The draft had not been read by the WG.  Ralph
will give the WG two weeks to read the draft (to begin 7/22), and then
start WG last call.

<draft-ietf-dhc-auth-suboption-00.txt>
--------------------------------------
Ralph pointed out the publication of "The Authentication Suboption
for the DHCP Relay Agent Option".  This draft is a result of a
discussion from the Minneapolis DHC WG meeting.  There are open issues
in the draft that will be discussed on the WG mailing list: hwo does a
device that doesn't set 'giaddr' get the response from the DHCP server
so the device cna remove its realy agent options; should the DHCP
specification be extended to explicitly allow chaining of relay agents
(as in DHCPv6)?

<draft-ietf-dhc-leasequery-03.txt>
----------------------------------
"DHCP Lease Query" <draft-ietf-dhc-leasequery-03.txt> is in WG last
call.  There have been some comments posted during the last call.
WG members were requested to further review the draft.

<draft-ietf-dhc-server-mib-06.txt>
----------------------------------
"Dynamic Host Configuration Protocol (DHCP) Server MIB"
<draft-ietf-dhc-server-mib-06.txt> is in WG last call.  There have
been some comments posted during the last call.  The authors will
respond to the comments and work with Rich Woundy to revise the
draft.

Q&A:
Narten: Deos this MIB have utility?
Droms: Every time Barr Hibbs has asked, WG consensus has been to
continue work on the MIB.
Narten: Does anyone plaen to implement this MIB?
Droms: Cisco does.
Woundy: CableLabs wants to build embedded DHCP servers in modems.
Some tables from this MIB would be useful.
Waters: Version -07 seems to have been lost; will work with Hibbs to
get versions straightened out.
Scanner: The -06 version is quite ISC-specific.
Droms: Please submit comments about that issue to the mailing list.

<draft-ietf-dhc-concat-04.txt>,  <draft-ietf-dhc-csr-07.txt>,
<draft-aboba-dhc-domsearch-09.txt>
-------------------------------------------------------------
The ADs have the token for review of "Encoding Long Options in DHCPv4"
<draft-ietf-dhc-concat-04.txt>, "The Classless Static Route Option for
DHCPv4" <draft-ietf-dhc-csr-07.txt> and "DHCP Domain Search Option"
<draft-aboba-dhc-domsearch-09.txt>.

<draft-ietf-dhc-dhcpv6-26.txt>
------------------------------
Ralph announced the status of the DHCPv6 spec: the -26 rev has been
reviewed by some IESG members and the authors are preparing a new rev
with responses to those reviews.  In particular, the WG was alerted to
changes in authentication and reconfiguration machinery.  Ralph asked
the WG to consider two issues: should the DHCPv6 authentication
mechanism be identified as a separate authentication protocol wihtin
the RFC3118 framework and should the "reconfiguration nonce" option
be merged into the RFC 3118 framework.  Some support was expressed for
the latter suggestion.

Ole Troan, <draft-troan-dhcpv6-opt-prefix-delegation-01.txt>
------------------------------------------------------------
Ole Troan gave an update presentation on "IPv6 Prefix Options for
DHCPv6" <draft-troan-dhcpv6-opt-prefix-delegation-01.txt> option.
Ralph mentioned that the IPv6 WG is working on a set of requirements
and a selectin process for prefix delegation.  There was some
discussion about modifications to the PD option to solve larger scale
problems - spec., make PD look more like Identity Association, with an
ID and similar semantics.  Authors will revise draft to accommodate
tat suggestion.

Other DHCPv6 option drafts
--------------------------
Ralph asked if the WG had reviewed the other DHCPv6 option drafts:
DNS Configuration options for DHCPv6
   <draft-ietf-dhc-dhcpv6-opt-dnsconfig-02.txt>
NIS Configuration Options for DHCPv6
   <draft-ietf-dhc-dhcpv6-opt-nisconfig-01.txt>
Time Configuration Options for DHCPv6
   <draft-ietf-dhc-dhcpv6-opt-timeconfig-01.txt>
DSTM Options for DHCP
   <draft-ietf-dhc-dhcpv6-opt-dstm-01.txt>
Client Preferred Prefix option for DHCPv6
   <draft-ietf-dhc-dhcpv6-opt-cliprefprefix-00.txt>
Some discussion of the use of DHCP for service configuration ensued.
Do we need a TTL field for service configuration options?  Other
server controls?  What's the problem to be solved? How does T1/T2 not
solve this problem today?  What's the overlap with the service
discovery protocol space?

DHC WG Charter
--------------
Ralph led a discussion of the WG charter.  He did some real-time edits
on the charter.  He will wordsmith the edits and continue the
discussion on the WG mailing list.

_______________________________________________
dhcwg mailing list
dhcwg@ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg