iFCP: minutes iFCP authors' confcall this Fri October 19th 9:00am PST

["Franco Travostino" <travos@nortelnetworks.com>]

Attendees:

Kevin Gibbons, Nishan Systems
Charles Monia, Nishan Systems
Inder Monga, Nortel Networks
Franco Travostino, Nortel Networks
Wayland Jeong,. Troika Networks

a) Comments on iFCP, -06.

CM described the delta between -05 and -06. The section on AES (with
proper I-D references) will be removed should AES hit any roadblock in
its way to RFC (to be verified at the upcoming Salt Lake City meeting,
IPsec WG meeting). CM mentioned that the mechanisms for comparing WWNs
(e.g., tie breaking) are currently under-specified and should rather
reference an RFC. All participants agreed that it will be good to receive
substantial review feedback based on -06 from the community.

b) iFCP Security Update 

FT described changes that occurred in the iFCP security words while they
were "massaged" within the security informational draft.
Changes are as follows:
1) "Conformant iFCP implementations MUST support ESP in tunnel mode
and MAY support ESP in transport mode" (the MAY was a SHOULD in
former iFCP text). 
2) "Manual keying MUST NOT be used". (missing in former iFCP
text)
3) Signature key authentication MAY be implemented (it was a SHOULD in
former iFCP text)
4) Aggressive mode SHOULD be used when pre-shared keys are used for
authentication ((it was a MUST in former iFCP text)
5) ID Payload MUST carry a single IP address and a single non-zero port
number (there wasn't a port number in former iFCP text). 
Changes 2-4 will be immediately retrofitted to the (authoritative) iFCP
specification text. There was consensus to wait for 1) and 5), with the
action item (IM the owner) to verify whether commercial, off the shelf
IKE implementations support this ID payload format.

CM queried about signature key authentication, which is still left as a
TBS in the iFCP spec. FT recalled that the authors of the security
informational draft are also waiting for new text on this topic. FT owns
the action item of checking with that crowd at their next confcall (Tue
23th).

c) iFCP MIB update

KG briefed the iFCP co-authors on the status of the iFCP MIB draft. The
next revision of the draft will take the official IETF name and will
restart with -00. As such, it must be turned in by the IETF 52 deadline
set for rookie drafts. Since Irvine,
Keith McCloghrie
<kzm@cisco.com>

has been very helpful. The MIB draft is known to compile without errors. The new draft will correctly cite rfc2837 for any Fibre Channel definition. In the new draft, there will also be a compliance section, and the gateway denominations will be removed from tables.

-franco

Franco Travostino, Director Content Internetworking Lab
Advanced Technology Investments
Nortel Networks, Inc.
600 Technology Park
Billerica, MA 01821 USA
Tel: 978 288 7708 Fax: 978 288 4690
email: travos@nortelnetworks.com