From majordomo@raleigh.ibm.com  Thu Jul  1 11:59:15 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA06959
	for <policy-archive@odin.ietf.org>; Thu, 1 Jul 1999 11:59:14 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id LAA20188;
	Thu, 1 Jul 1999 11:57:17 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id LAA31944;
	Thu, 1 Jul 1999 11:57:06 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA149986; Thu, 1 Jul 1999 11:39:33 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA223450; Thu, 1 Jul 1999 11:39:28 -0400
Received: from southrelay02.raleigh.ibm.com (southrelay02.raleigh.ibm.com [9.37.3.209])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id LAA34816
	for <policy@raleigh.ibm.com>; Thu, 1 Jul 1999 11:39:33 -0400
From: remoore@us.ibm.com
Received: from d54mta04.raleigh.ibm.com (d54mta04.raleigh.ibm.com [9.67.228.36])
	by southrelay02.raleigh.ibm.com (8.8.8m2/NCO v2.03) with SMTP id LAA50708;
	Thu, 1 Jul 1999 11:39:32 -0400
Received: by d54mta04.raleigh.ibm.com(Lotus SMTP MTA v4.6.4  (830.2 3-23-1999))  id 852567A1.00560221 ; Thu, 1 Jul 1999 11:39:26 -0400
X-Lotus-Fromdomain: IBMUS
To: policy@raleigh.ibm.com
Cc: dcook@cabletron.com
Message-Id: <852567A1.0056001E.00@d54mta04.raleigh.ibm.com>
Date: Thu, 1 Jul 1999 11:30:43 -0400
Subject: Fwd: ietf core policy schema draft4 question
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: remoore@us.ibm.com



Dana is absolutely right -- some of the class definitions in the
-04 LDAP Core Policy Schema draft don't match the class hierarchy
described in Figure 1 and in the overall text of the document.
Looking back, though, I find that these errors were not introduced
in the -04 version:  they were also present in the -03 document.

In any case, here are the corrections I've made in the document
source, in anticipation of eventually publishing an -05 version:

Section 5.2 (The Class policyGroup)

    change
      DERIVED FROM     top
    to
      DERIVED FROM     policy


Section 5.3 (The Class policyRule)

    change
      DERIVED FROM     top
    to
      DERIVED FROM     policy

    change
      AUXILIARY CLASSES    none
    to
     AUXILIARY CLASSES    policyCondition, policyAction


Section 5.4 (The Class policyCondition)

    change
      DERIVED FROM     top
    to
      DERIVED FROM     policy


Section 5.7 (The Class policyAction)

    change
      DERIVED FROM     top
    to
      DERIVED FROM     policy


Regards,
Bob

Bob Moore
IBM Networking Software
+1-919-254-4436
remoore@us.ibm.com


---------------------- Forwarded by Robert Moore/Raleigh/IBM on 07/01/99 11:03
AM ---------------------------
>Date: Tue, 29 Jun 1999 17:11:16 -0400
>X-Mailer: Mozilla 4.6 [en] (X11; U; SunOS 5.7 sun4u)
>X-Accept-Language: en
>To: johns@cisco.com, ellesson@raleigh.ibm.com
>CC: "Bucknam, Steven" <bucknam@cabletron.com>,
>        "Calato, Paul" <calato@cabletron.com>
>Subject: ietf core policy schema draft4 question
>
>
>John or Ed,
>
>    I am confused about a point in the draft4 schema that I hope is
>just a typo. Your class heirarchy diagram and the text seem to suggest
>that core policy objects derive from "policy", but policyRule
>is specified to derive from top. Shouldn't it derive from policy, or
>mix in policyElement to get the policyKeywords attribute?
>
>Also, shouldn't policyCondition and policyAction be listed as possible
>auxiliary classes to policy rule ( for creating simple policies ).
>
>
>                   Thanks,
>
>                        Dana
>
>
>--
>Dana P. Cook
>Cabletron Systems Inc.
>(603) 337-7551
>dcook@ctron.com
>




From majordomo@raleigh.ibm.com  Thu Jul  1 12:35:22 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA08484
	for <policy-archive@odin.ietf.org>; Thu, 1 Jul 1999 12:35:21 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA21226;
	Thu, 1 Jul 1999 12:24:44 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id MAA28664;
	Thu, 1 Jul 1999 12:24:28 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA97666; Thu, 1 Jul 1999 12:13:35 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA149822; Thu, 1 Jul 1999 12:13:21 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id MAA32036
	for <policy@raleigh.ibm.com>; Thu, 1 Jul 1999 12:13:21 -0400
Received: from ckmso1.proxy.att.com (ckmso1.att.com [12.20.58.69])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA13448
	for <policy@raleigh.ibm.com>; Thu, 1 Jul 1999 12:13:13 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by ckmso1.proxy.att.com (AT&T IPNS/MS-2.2) with SMTP id MAA18153;
	Thu, 1 Jul 1999 12:12:05 -0400 (EDT)
Received: by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id MAA01420; Thu, 1 Jul 1999 12:12:01 -0400
Date: Thu, 1 Jul 1999 12:12:01 -0400
Message-Id: <199907011612.MAA01420@qsun.ho.att.com>
From: jayhawk@qsun.ho.att.com (Ryan Moats)
To: policy@raleigh.ibm.com
Cc: remoore@us.ibm.com
Subject: Initial comments on draft-ietf-policy-core-schema-04.txt
Content-Type: text
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: jayhawk@qsun.ho.att.com (Ryan Moats)

I haven't seen this, so I'm resubmitting on the basis
of a mail failure.  My apologies to those who see this
twice...

| I've submitted the two Core Policy Internet-Drafts (i.e., the
| information model and the LDAP representation of it) to the
| Internet-Drafts repository.  The drafts are also available on
| our AIW FTP site, at the following URLs:
| 
| ftp://ftp.networking.ibm.com/pub/standards/aiw/misc
|      /draft-ietf-policy-core-info-model-00.txt
| 
| ftp://ftp.networking.ibm.com/pub/standards/aiw/misc
|      /draft-ietf-policy-core-schema-04.txt
| 

Having started to read the schema draft, I'm
bothered that it does not follow the ABNF specified
in RFC 2252, which (at least for me) makes it a
little more difficult to follow than otherwise.
Are there any plans to redo the schema so that it conforms?

Ryan



From majordomo@raleigh.ibm.com  Mon Jul  5 13:49:14 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA03874
	for <policy-archive@odin.ietf.org>; Mon, 5 Jul 1999 13:49:13 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA23038;
	Mon, 5 Jul 1999 13:49:14 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id NAA24692;
	Mon, 5 Jul 1999 13:49:11 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA44846; Mon, 5 Jul 1999 13:40:36 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA35860; Mon, 5 Jul 1999 13:40:29 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id NAA05716;
	Mon, 5 Jul 1999 13:40:31 -0400
Received: from nsm-mail2.cisco.com (nsm-mail2.cisco.com [171.71.236.25])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA21312;
	Mon, 5 Jul 1999 13:40:28 -0400
Received: from jstrassnlap.cisco.com ([171.69.108.130])
	by nsm-mail2.cisco.com (8.8.8-Cisco List Logging/8.8.8) with SMTP id KAA20936;
	Mon, 5 Jul 1999 10:39:57 -0700 (PDT)
Message-Id: <02f201bec70d$a33def40$826c45ab@cisco.com>
From: "John Strassner" <jstrassn@cisco.com>
To: <policy@raleigh.ibm.com>
Cc: <ysnir@cisco.com>, <yramberg@cisco.com>,
        "Ed Ellesson" <ellesson@raleigh.ibm.com>
Subject: QoS Policy schema draft
Date: Mon, 5 Jul 1999 10:41:39 -0700
Organization: Cisco Systems
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "John Strassner" <jstrassn@cisco.com>
Content-Transfer-Encoding: 7bit

Policy Enthusiasts,

The following draft was submitted to the Policy Framework working
group as an Internet Draft, but appears to have missed the submission
cutoff date:
  draft-snir-policy-qos-infomodel-00.txt

In the meantime, it may be accessed by anonymous FTP from:
  ftp://ftpeng.cisco.com/ftp/public/policy

This document presents a Quality of Service (QoS) Policy schema,
based on the concepts and classes defined in the Policy Framework
Core Information Model and Schema documents. The purpose of
this document is to present a means to model network (QoS)
policy information. This information is typically used by QoS Policy
Servers to configure network devices according to prescribed QoS
Policies.

regards,
John



From majordomo@raleigh.ibm.com  Wed Jul  7 14:04:19 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA16842
	for <policy-archive@odin.ietf.org>; Wed, 7 Jul 1999 14:04:18 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id OAA30726;
	Wed, 7 Jul 1999 14:04:15 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id OAA23530;
	Wed, 7 Jul 1999 14:03:53 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA47584; Wed, 7 Jul 1999 13:46:45 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA80328; Wed, 7 Jul 1999 13:46:41 -0400
Received: from ellesson.raleigh.ibm.com (ellesson.raleigh.ibm.com [9.37.84.214])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id NAA34950
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 13:46:45 -0400
Message-Id: <199907071746.NAA34950@rtpmail02.raleigh.ibm.com>
X-Sender: ellesson@dnjake2.raleigh.ibm.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0
Date: Wed, 07 Jul 1999 13:47:19 -0400
To: policy@raleigh.ibm.com
From: Ed Ellesson <ellesson@raleigh.ibm.com>
Subject: Fwd: Tentative Agenda for Policy Framework WG Meeting in Oslo
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Ed Ellesson <ellesson@raleigh.ibm.com>

Hello, 

It appears that I sent this to the agenda for posting on the Oslo meeting
web page, but neglected to send it to the list.   Here it is.  Those of you
who are tentatively on the agenda should already have seen this.  
 
>Date: Thu, 01 Jul 1999 09:46:34 -0400
>To: agenda@ietf.org
>From: Ed Ellesson <ellesson@raleigh.ibm.com>
>Subject: Tentative Agenda for Policy Framework WG Meeting in Oslo
>
>Tentative Agenda for Policy Framework WG Meeting in Oslo
>
>Monday, 1930 - 2200
>-------------------	
>
>1. Intro, Objectives, Agenda Bash, Charter Status (John Strassner) (20 min.)
>
>2. Core Information Model for Policy (Bob Moore) (20 min)
>
>3. LDAP-Accessible Directory Schema for Policy (Bob Moore) (40 min)
>
>4. QOS Policy Approaches (20 minutes each x 3 = 60 min total)
>
>   -Raju Rajan 
>   -Walter Weiss, et al
>   -Yoram Snir
>
>5. Summary, first day (John Strassner) (10 min)
>
>Tuesday, 1545 - 1800
>--------------------
>
>6. Intro to second day (Ed Ellesson) (10 min)
>
>5.Architecture Draft Issues (Glenn Waters ? ) (30 min)
>
>6. Other Working Groups (20 min each = ?)
>
>   -Presentation on PIB's (Andrew, Keith, or Scott?)
>   -Other related work?  
>
>7. Usage Scenarios (Hugh Mahon ? ) (25 min)
>
>8. Summary (Ed Ellesson)  (30 min)
>
>   -Summary of meeting
>   -Working group last calls to be scheduled
>   -Next Steps 

Regards, 

Ed Ellesson
Sr. Engineer, TCP/IP Technology Management
IBM Networking Software Products
Research Triangle Park, NC 
ellesson@raleigh.ibm.com
919-254-4115


From majordomo@raleigh.ibm.com  Wed Jul  7 15:28:47 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id PAA21519
	for <policy-archive@odin.ietf.org>; Wed, 7 Jul 1999 15:28:45 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id PAA24720;
	Wed, 7 Jul 1999 15:27:58 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id PAA22720;
	Wed, 7 Jul 1999 15:27:49 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA32722; Wed, 7 Jul 1999 15:14:49 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA36544; Wed, 7 Jul 1999 15:14:44 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id PAA28584;
	Wed, 7 Jul 1999 15:14:50 -0400
Received: from bmailnj.iphighway.com ([209.3.6.76])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id PAA23838;
	Wed, 7 Jul 1999 15:14:46 -0400
Received: by BMAILNJ with Internet Mail Service (5.5.2448.0)
	id <3183YXQF>; Wed, 7 Jul 1999 15:07:47 -0400
Message-Id: <6399122981E1D211AB490090271E0AA31333B1@BMAILNJ>
From: Francis Reichmeyer -NJ <FranR@iphighway.com>
To: "'Ed Ellesson'" <ellesson@raleigh.ibm.com>, policy@raleigh.ibm.com
Subject: RE: Tentative Agenda for Policy Framework WG Meeting in Oslo
Date: Wed, 7 Jul 1999 15:07:46 -0400 
Mime-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Francis Reichmeyer -NJ <FranR@iphighway.com>

What's going to be discussed in this segment? Is there a new draft?

Thanks,
-Fran


> -----Original Message-----
> From: Ed Ellesson [mailto:ellesson@raleigh.ibm.com]
> Sent: Wednesday, July 07, 1999 1:47 PM
> To: policy@raleigh.ibm.com
> Subject: Fwd: Tentative Agenda for Policy Framework WG Meeting in Oslo
> 
> 
> >
> >5.Architecture Draft Issues (Glenn Waters ? ) (30 min)
> >


From majordomo@raleigh.ibm.com  Wed Jul  7 17:00:38 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA23216
	for <policy-archive@odin.ietf.org>; Wed, 7 Jul 1999 17:00:37 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id RAA31232;
	Wed, 7 Jul 1999 17:00:37 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id RAA29896;
	Wed, 7 Jul 1999 17:00:32 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA22642; Wed, 7 Jul 1999 16:53:23 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA22622; Wed, 7 Jul 1999 16:53:17 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id QAA32932
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 16:53:22 -0400
Received: from apollo.sitaranetworks.com (apollo.sitaranetworks.com [199.103.141.105])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id QAA27130
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 16:53:17 -0400
Message-Id: <01a501bec8bb$2e3027e0$728d67c7@sitaranetworks.com>
From: "Vladimir Sukonnik" <vladimir@sitaranetworks.com>
To: <policy@raleigh.ibm.com>
Subject: how scalable is the policy framework?
Date: Wed, 7 Jul 1999 16:56:26 -0400
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Vladimir Sukonnik" <vladimir@sitaranetworks.com>
Content-Transfer-Encoding: 7bit

Hello,

I've read the policy framework and COPS internet drafts and have a basic
question. I hope you can help me to clear up
my confusion. I don't see how this framework would scale for thousands for
PEPs. According to these documents certain events
trigger PEPs into calling up PDPs for "the final policy decision". Wouldn't
this make the PDP a potential bottleneck and
slow PEPs down to a crawl while they wait for that decision to come down?
What am I missing?

Thank you,
Vladimir.



From majordomo@raleigh.ibm.com  Wed Jul  7 17:34:09 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA23884
	for <policy-archive@odin.ietf.org>; Wed, 7 Jul 1999 17:34:08 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id RAA24232;
	Wed, 7 Jul 1999 17:33:42 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id RAA35298;
	Wed, 7 Jul 1999 17:32:46 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA77926; Wed, 7 Jul 1999 17:22:41 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA77406; Wed, 7 Jul 1999 17:22:38 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id RAA37398
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 17:22:38 -0400
Received: from thalia.fm.intel.com (thalia.fm.intel.com [132.233.247.11])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id RAA22112
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 17:22:35 -0400
Received: from fmsmsx19.fm.intel.com (fmsmsx19.fm.intel.com [132.233.222.210])
	by thalia.fm.intel.com (8.9.1a+p1/8.9.1/d: relay.m4,v 1.6 1998/11/24 22:10:56 iwep Exp iwep $) with ESMTP id OAA08149;
	Wed, 7 Jul 1999 14:22:29 -0700 (PDT)
Received: by fmsmsx19.fm.intel.com with Internet Mail Service (5.5.2448.0)
	id <3N68BL77>; Wed, 7 Jul 1999 14:22:29 -0700
Message-Id: <4A043A1FE4B2D111AC3F00A0C96B513301B0F637@fmsmsx37.fm.intel.com>
From: "Durham, David" <david.durham@intel.com>
To: "'Vladimir Sukonnik'" <vladimir@sitaranetworks.com>,
        policy@raleigh.ibm.com
Subject: RE: how scalable is the policy framework?
Date: Wed, 7 Jul 1999 14:22:26 -0700 
Mime-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0)
Content-Type: text/plain
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Durham, David" <david.durham@intel.com>

From the perspective of the RAP WG, COPS can be used so 
a PEP can outsource its questions to a PDP when it does
not know better. This mechanism is useful for coarse granularity
events, such as the arrival of an RSVP reservation or 
authentication request. Likewise, the PDP can use COPS to 
dynamically and continuously configure simple access control 
lists on its PEPs such that the PEPs can make simple policy 
decisions locally (based on 5-tuple filters and device specific
actions). The latter is ongoing work related to COPS-PR and the 
definition of the PIB (Policy Information Base).

Thus, COPS achieves a balance between scalability and flexibility
all within a notification-based architecture that alleviates the 
need for continuos polling. You may also note that there will 
likely be several PDPs within a domain containing thousands of 
PEPs, each responsible for only a subset of the PEPs.

-Dave


> -----Original Message-----
> From:	Vladimir Sukonnik [SMTP:vladimir@sitaranetworks.com]
> Sent:	Wednesday, July 07, 1999 1:56 PM
> To:	policy@raleigh.ibm.com
> Subject:	how scalable is the policy framework?
> 
> Hello,
> 
> I've read the policy framework and COPS internet drafts and have a basic
> question. I hope you can help me to clear up
> my confusion. I don't see how this framework would scale for thousands for
> PEPs. According to these documents certain events
> trigger PEPs into calling up PDPs for "the final policy decision".
> Wouldn't
> this make the PDP a potential bottleneck and
> slow PEPs down to a crawl while they wait for that decision to come down?
> What am I missing?
> 
> Thank you,
> Vladimir.


From majordomo@raleigh.ibm.com  Wed Jul  7 18:29:17 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA24389
	for <policy-archive@odin.ietf.org>; Wed, 7 Jul 1999 18:29:16 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id SAA22518;
	Wed, 7 Jul 1999 18:29:10 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id SAA33254;
	Wed, 7 Jul 1999 18:27:58 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA44022; Wed, 7 Jul 1999 18:20:24 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA39644; Wed, 7 Jul 1999 18:20:17 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id SAA29494
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 18:20:16 -0400
Received: from mail-blue.research.att.com (mail-blue.research.att.com [135.207.30.102])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id SAA16336
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 18:20:12 -0400
Received: from bigmail.research.att.com (bigmail.research.att.com [135.207.30.101])
	by mail-blue.research.att.com (Postfix) with ESMTP
	id 0FD144CE08; Wed,  7 Jul 1999 18:20:15 -0400 (EDT)
Received: from pcraju (pc-raju.research.att.com [135.207.27.140])
	by bigmail.research.att.com (8.8.8/8.8.8) with SMTP id SAA27827;
	Wed, 7 Jul 1999 18:20:14 -0400 (EDT)
Message-Id: <022601bec8c8$05420620$8c1bcf87@research.att.com>
From: "Raju Rajan" <rajan@research.att.com>
To: "Durham, David" <david.durham@intel.com>,
        "'Vladimir Sukonnik'" <vladimir@sitaranetworks.com>,
        <policy@raleigh.ibm.com>
References: <4A043A1FE4B2D111AC3F00A0C96B513301B0F637@fmsmsx37.fm.intel.com>
Subject: Re: how scalable is the policy framework?
Date: Wed, 7 Jul 1999 18:28:21 -0400
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Raju Rajan" <rajan@research.att.com>
Content-Transfer-Encoding: 7bit

Vladimir,

Scalability in just the size of the network is a partial view; one must
consider other factors including:
1) What is the operational environment? Eg. LANs, ISP backbone, cable modem
banks, phone modems, wireless, etc
2) What are the services you are trying to provide? VPNs, pay-per-view,
content filtering, remote access, privacy, etc.
3) What is the "fragmentation" of your policy base? i.e., How many users,
devices, accounts, application categories, administrative domains, etc.
4) What is the dynamism of the environment? ie., how many flows, what is the
rate of rule change, binding services turnover (think DHCP, for eg.) etc.

In other words, even LDAP alone (with replication) can scale to 1000s of
devices if you have relatively static, simple and stateless rules. A
reasonable solution like COPS+LDAP+SNMP+Radius/Diameter (deployed in
different parts of your network) can cover much more ground. The answer to
scalability can only be given when you pose the question more precisely.
(Anyway, I suspect it will be "not right now!" :)

IMHO, the real problem is not the number of devices; at least 90% of the
1000s of PEPs you're talking about will be policy unaware for the next 4-5
years. The question is what are you going to do about them?

raju rajan


----- Original Message -----
From: Durham, David <david.durham@intel.com>
To: 'Vladimir Sukonnik' <vladimir@sitaranetworks.com>;
<policy@raleigh.ibm.com>
Sent: Wednesday, July 07, 1999 5:22 PM
Subject: RE: how scalable is the policy framework?


From the perspective of the RAP WG, COPS can be used so
a PEP can outsource its questions to a PDP when it does
not know better. This mechanism is useful for coarse granularity
events, such as the arrival of an RSVP reservation or
authentication request. Likewise, the PDP can use COPS to
dynamically and continuously configure simple access control
lists on its PEPs such that the PEPs can make simple policy
decisions locally (based on 5-tuple filters and device specific
actions). The latter is ongoing work related to COPS-PR and the
definition of the PIB (Policy Information Base).

Thus, COPS achieves a balance between scalability and flexibility
all within a notification-based architecture that alleviates the
need for continuos polling. You may also note that there will
likely be several PDPs within a domain containing thousands of
PEPs, each responsible for only a subset of the PEPs.

-Dave


> -----Original Message-----
> From: Vladimir Sukonnik [SMTP:vladimir@sitaranetworks.com]
> Sent: Wednesday, July 07, 1999 1:56 PM
> To: policy@raleigh.ibm.com
> Subject: how scalable is the policy framework?
>
> Hello,
>
> I've read the policy framework and COPS internet drafts and have a basic
> question. I hope you can help me to clear up
> my confusion. I don't see how this framework would scale for thousands for
> PEPs. According to these documents certain events
> trigger PEPs into calling up PDPs for "the final policy decision".
> Wouldn't
> this make the PDP a potential bottleneck and
> slow PEPs down to a crawl while they wait for that decision to come down?
> What am I missing?
>
> Thank you,
> Vladimir.




From majordomo@raleigh.ibm.com  Wed Jul  7 18:32:20 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA24474
	for <policy-archive@odin.ietf.org>; Wed, 7 Jul 1999 18:32:19 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id SAA34092;
	Wed, 7 Jul 1999 18:32:10 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id SAA36834;
	Wed, 7 Jul 1999 18:29:59 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA84052; Wed, 7 Jul 1999 18:25:47 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA37452; Wed, 7 Jul 1999 18:25:44 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id SAA37188
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 18:25:44 -0400
Received: from bmailnj.iphighway.com ([209.3.6.76])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id SAA26958
	for <policy@raleigh.ibm.com>; Wed, 7 Jul 1999 18:25:40 -0400
Received: from shai-notebook (SHAI_NOTEBOOK [209.3.6.106]) by bmailnj.iphighway.com with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2448.0)
	id 3183YXRX; Wed, 7 Jul 1999 18:18:51 -0400
Message-Id: <4.2.0.58.19990707180117.00a5af00@bmailnj.iphighway.com>
X-Sender: herzog@bmailnj.iphighway.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.0.58 
Date: Wed, 07 Jul 1999 18:20:27 -0400
To: "Vladimir Sukonnik" <vladimir@sitaranetworks.com>,
        <policy@raleigh.ibm.com>, rap@iphighway.com
From: Shai Herzog <herzog@iphighway.com>
Subject: Re: how scalable is the policy framework?
In-Reply-To: <01a501bec8bb$2e3027e0$728d67c7@sitaranetworks.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Shai Herzog <herzog@iphighway.com>

Vladimir,

You are probably correct in your assumption that a single PDP wouldn't
be able to scale to thousands of ACTIVE PEPs. Fortunately, it has been
shown that policy enforcement is really a distributed process.

On one extreme, one PDP controls an entire huge network. On the other
extreme, each device/client fends for itself and makes its own policy
decisions. In the latter case, local decisions are at risk of being
uncoordinated with the rest of the network.

The framework allows a flexible adoption of those two extremes as
well as the continuum in between. It allows some devices to have both
PDP and PEP internally (to be free of management by any external PDP).
It also allows multiple PDPs to reside in the same domain, where each
PDP is responsible for a distinct set of devices.

So, I believe the framework doesn't restrict scalability. In fact,
the distributed solution has very good scaling properties and can scale
to virtually any size network.

Shai

P.S., As an example you can look at our OPS system, which employs
a fully distributed architecture with multiple PDPs, all
coordinated under the same management infrastructure.

At 04:56 PM 7/7/99, Vladimir Sukonnik wrote:
>Hello,
>
>I've read the policy framework and COPS internet drafts and have a basic
>question. I hope you can help me to clear up
>my confusion. I don't see how this framework would scale for thousands for
>PEPs. According to these documents certain events
>trigger PEPs into calling up PDPs for "the final policy decision". Wouldn't
>this make the PDP a potential bottleneck and
>slow PEPs down to a crawl while they wait for that decision to come down?
>What am I missing?
>
>Thank you,
>Vladimir.


__________________________________________________________
Shai Herzog                  Founder & CTO, IPHighway Inc.
Cell: (408) 390-3045       http://www.iphighway.com/herzog
Tel : (201) 585-0800                  Fax : (212) 656-1006 


From majordomo@raleigh.ibm.com  Sat Jul 10 07:10:58 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA08877
	for <policy-archive@odin.ietf.org>; Sat, 10 Jul 1999 07:10:53 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id HAA23584;
	Sat, 10 Jul 1999 07:10:33 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id HAA34796;
	Sat, 10 Jul 1999 07:10:28 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA80362; Sat, 10 Jul 1999 06:52:58 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA31954; Sat, 10 Jul 1999 06:52:54 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id GAA29822
	for <policy@raleigh.ibm.com>; Sat, 10 Jul 1999 06:52:58 -0400
Received: from omega.cisco.com (omega.cisco.com [171.69.63.141])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id GAA23734
	for <policy@raleigh.ibm.com>; Sat, 10 Jul 1999 06:52:56 -0400
Received: from jstrassnlap.cisco.com (sj-dial-3-232.cisco.com [171.68.180.233])
	by omega.cisco.com (8.8.8-Cisco List Logging/8.8.8) with SMTP id DAA20180
	for <policy@raleigh.ibm.com>; Sat, 10 Jul 1999 03:52:24 -0700 (PDT)
Message-Id: <003c01becac2$90f69d60$e9b444ab@cisco.com>
From: "John Strassner" <jstrassn@cisco.com>
To: <policy@raleigh.ibm.com>
Subject: QoS Drafts
Date: Sat, 10 Jul 1999 03:49:39 -0700
Organization: Cisco Systems
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "John Strassner" <jstrassn@cisco.com>
Content-Transfer-Encoding: 7bit

Policy Enthusiasts,

some of you have asked for the names of the different QoS drafts that
will be discussed in Oslo, which is item #4 in the agenda - here they
are:

>4. QOS Policy Approaches (20 minutes each x 3 = 60 min total)
>
>   -Raju Rajan
        draft-rajan-policy-conditions-00.txt
        draft-rajan-policy-qosschema.txt
>   -Walter Weiss, et al
        draft-weiss-policy-device-qos-model-00.txt
>   -Yoram Snir
        draft-snir-policy-qos-infomodel-00.txt
        (this can be pulled from
ftp://ftpeng.cisco.com/ftp/public/policy)

regards,
John



From majordomo@raleigh.ibm.com  Fri Jul 16 16:28:19 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA09316
	for <policy-archive@odin.ietf.org>; Fri, 16 Jul 1999 16:28:18 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id QAA38902;
	Fri, 16 Jul 1999 16:28:18 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id QAA23058;
	Fri, 16 Jul 1999 16:27:53 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA143578; Fri, 16 Jul 1999 16:08:14 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA54992; Fri, 16 Jul 1999 16:08:11 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id QAA26632
	for <policy@raleigh.ibm.com>; Fri, 16 Jul 1999 16:08:12 -0400
Received: from dirty.research.bell-labs.com (dirty.research.bell-labs.com [204.178.16.6])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with SMTP id QAA33106
	for <policy@raleigh.ibm.com>; Fri, 16 Jul 1999 16:08:08 -0400
Received: from scummy.research.bell-labs.com ([135.104.2.10]) by dirty; Fri Jul 16 16:06:56 EDT 1999
Received: from zydeco.research.bell-labs.com ([135.104.120.150]) by scummy; Fri Jul 16 16:07:01 EDT 1999
Received: from research.bell-labs.com (summer118-02.research.bell-labs.com [135.104.121.21])
	by zydeco.research.bell-labs.com (8.9.1/8.9.1) with ESMTP id QAA16017
	for <policy@raleigh.ibm.com>; Fri, 16 Jul 1999 16:07:00 -0400 (EDT)
Message-Id: <378F9F49.B4936427@research.bell-labs.com>
Date: Fri, 16 Jul 1999 14:08:25 -0700
From: Pradosh Kumar Mohapatra <mohap@research.bell-labs.com>
X-Mailer: Mozilla 4.5 [en] (WinNT; U)
X-Accept-Language: en
Mime-Version: 1.0
To: policy <policy@raleigh.ibm.com>
Subject: COPS message content
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Pradosh Kumar Mohapatra <mohap@research.bell-labs.com>
Content-Transfer-Encoding: 7bit

Hi,
Being new to the mailing-list, I really don't know whether there is any
information available on the message content of a COPS message sent from
PEP to PDP or PDP to PEP. I am referring to the ClientSI part of the
COPS messages. Can't there be any standard way to represent this part of
the protocol, given the fact that we have the ldap structure of policy
in place.
From whatever I have read, I came to understand that this field of the
protocol is very much specific to the discipline that it related to e.g.
qos, or security. If it is so, then after the packet reaches the PDP,
how does the PDP operate? What are the details about how it accesses the
LDAP and matches a rule? If a rule is matched, what contents does it put
in the reply message?

Regards,
Pradosh



From majordomo@raleigh.ibm.com  Mon Jul 19 19:03:11 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id TAA04219
	for <policy-archive@odin.ietf.org>; Mon, 19 Jul 1999 19:03:10 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id TAA35292;
	Mon, 19 Jul 1999 19:03:09 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id TAA34308;
	Mon, 19 Jul 1999 19:02:59 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA50000; Mon, 19 Jul 1999 18:47:25 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA51014; Mon, 19 Jul 1999 18:47:22 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id SAA30278
	for <policy@raleigh.ibm.com>; Mon, 19 Jul 1999 18:47:22 -0400
Received: from sol.extremenetworks.com (sol.extremenetworks.com [216.52.8.2])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id SAA27348
	for <policy@raleigh.ibm.com>; Mon, 19 Jul 1999 18:47:18 -0400
Received: by ftp.extremenetworks.com with Internet Mail Service (5.5.2448.0)
	id <36G72R94>; Mon, 19 Jul 1999 15:46:50 -0700
Message-Id: <D0805D3B448BD211A7990008C7B1813014306F@ftp.extremenetworks.com>
From: Andrew Smith <andrew@extremenetworks.com>
To: "'Pradosh Kumar Mohapatra'" <mohap@research.bell-labs.com>
Cc: "rap (E-mail)" <rap@iphighway.com>, policy <policy@raleigh.ibm.com>
Subject: RE: COPS message content
Date: Mon, 19 Jul 1999 15:46:46 -0700
Mime-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Andrew Smith <andrew@extremenetworks.com>

Pradosh,

This is partly a question for the RAP WG list.

The base COPS protocol
(http://www.ietf.org/internet-drafts/draft-ietf-rap-cops-06.txt) does not
define the contents for the ClientSI objects.

So far the only COPS client-type fully defined is that for use with RSVP.
The contents of the ClientSI objects in this application are defined in
section 2.3 of
http://www.ietf.org/internet-drafts/draft-ietf-rap-cops-rsvp-05.txt. This
section includes by reference many of the object formats from the RSVP
protocol itself (RFC2205). It did not make sense to try to force these
existing RSVP objects into a different common format for this application.

For the "COPS for Provisioning" application, there is a recent proposal for
the contents of the ClientSI objects in
http://www.ietf.org/internet-drafts/draft-ietf-rap-pr-00.txt. As you
correctly point out, this does not currently reflect any of the
syntax/organisation of LDAP schemas from which the provisioning rules might
be derived. The working assumption (I think) is that a PDP would have to
take rules out of a "policy repository" (this might be an LDAP directory or
it might also be some other type of store) and translate them in some way
before provisioning the appropriate rules into a PEP. It is not clear that
there is always a suitable direct translation into PEP-speak - more likely
there is quite a complex algorithm to be executed by the PDP in taking
dynamic network conditions into consideration (e.g. topology, time-of-day,
traffic patterns etc.) in this algorithm. So, IMHO, a tight coupling of COPS
objects to LDAP syntax would seem like the wrong approach.

I think that the details of the translation algorithm run by the PDP can,
and will, be left for implementor differentiation: only the on-the-wire
formats for COPS (and LDAP) need to be standardised although we will
probably need to discuss the internal PDP algorithms in this working group
to some extent in order to make sure that the wire-protocols are adequate
for the job (this is the proof-of-concept stuff that is talked about in the
Policy WG charter).

I hope that is a fair summary of the current consensus.


Andrew Smith
RAP WG co-chair


> -----Original Message-----
> From: Pradosh Kumar Mohapatra [mailto:mohap@research.bell-labs.com]
> Sent: Friday, July 16, 1999 2:08 PM
> To: policy
> Subject: COPS message content
> 
> 
> Hi,
> Being new to the mailing-list, I really don't know whether 
> there is any
> information available on the message content of a COPS 
> message sent from
> PEP to PDP or PDP to PEP. I am referring to the ClientSI part of the
> COPS messages. Can't there be any standard way to represent 
> this part of
> the protocol, given the fact that we have the ldap structure of policy
> in place.
> From whatever I have read, I came to understand that this field of the
> protocol is very much specific to the discipline that it 
> related to e.g.
> qos, or security. If it is so, then after the packet reaches the PDP,
> how does the PDP operate? What are the details about how it 
> accesses the
> LDAP and matches a rule? If a rule is matched, what contents 
> does it put
> in the reply message?
> 
> Regards,
> Pradosh
> 


From majordomo@raleigh.ibm.com  Mon Jul 19 23:19:16 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id XAA07864
	for <policy-archive@odin.ietf.org>; Mon, 19 Jul 1999 23:19:16 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id XAA22464;
	Mon, 19 Jul 1999 23:19:12 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id XAA27876;
	Mon, 19 Jul 1999 23:18:55 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA42602; Mon, 19 Jul 1999 23:04:34 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA32866; Mon, 19 Jul 1999 23:04:31 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id XAA34680
	for <policy@raleigh.ibm.com>; Mon, 19 Jul 1999 23:04:36 -0400
Received: from mail.webstream.net (mail.webstream.net [209.203.197.7])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id XAA27546
	for <policy@raleigh.ibm.com>; Mon, 19 Jul 1999 23:04:31 -0400
Received: from deterministic ([205.179.103.218]) by mail.webstream.net
          (Post.Office MTA v3.1 release PO205e ID# 0-12345L500S10000V35)
          with SMTP id AAA261; Mon, 19 Jul 1999 23:06:12 -0400
Message-Id: <001501bed25c$fa74dee0$da67b3cd@DeterministicNetworks.com>
From: stevej@DeterministicNetworks.com (SteveJ)
To: "Andrew Smith" <andrew@extremenetworks.com>,
        "'Pradosh Kumar Mohapatra'" <mohap@research.bell-labs.com>
Cc: "rap (E-mail)" <rap@iphighway.com>, "policy" <policy@raleigh.ibm.com>
References: <D0805D3B448BD211A7990008C7B1813014306F@ftp.extremenetworks.com>
Subject: Re: COPS message content
Date: Mon, 19 Jul 1999 20:07:17 -0700
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2014.211
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2014.211
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: stevej@DeterministicNetworks.com (SteveJ)
Content-Transfer-Encoding: 7bit



Pradosh and Andrew,

>
> For the "COPS for Provisioning" application, there is a recent proposal
for
> the contents of the ClientSI objects in
> http://www.ietf.org/internet-drafts/draft-ietf-rap-pr-00.txt. As you
> correctly point out, this does not currently reflect any of the
> syntax/organisation of LDAP schemas from which the provisioning rules
might
> be derived. The working assumption (I think) is that a PDP would have to
> take rules out of a "policy repository" (this might be an LDAP directory
or
> it might also be some other type of store) and translate them in some way
> before provisioning the appropriate rules into a PEP. It is not clear that
> there is always a suitable direct translation into PEP-speak - more likely
> there is quite a complex algorithm to be executed by the PDP in taking
> dynamic network conditions into consideration (e.g. topology, time-of-day,
> traffic patterns etc.) in this algorithm. So, IMHO, a tight coupling of
COPS
> objects to LDAP syntax would seem like the wrong approach.
>
While I agree that a tight coupling of LDAP to COPS would be the wrong
approach, I can't help thinking that the policy objects themselves should
more closely follow what is described in the Policy Framework.  I believe
the jury is still out on how the Policy Server/Repository -- PDP -- PEP will
actually be implemented.  Today there are vendors who think these are three
separate entities, others who think the Policy Server and PDP are combined,
and still others who think the PDP and PEP should be combined.  In my
personal opinion, the first has scaling problems with per-flow requests to
the PDP.  The second has scaling problems with PEP requests to the Policy
Server, and only the third seems uniformly scalable - policies and
associated rules are distributed to remote PDP/PEP entities from the
repository.  The distribution may take multiple forms - LDAP for
bulk/default policies, COPS, SNMP or other protocol for realtime
updates/changes based on network conditions that can't be defined in the
initial ruleset.

> I think that the details of the translation algorithm run by the PDP can,
> and will, be left for implementor differentiation: only the on-the-wire
> formats for COPS (and LDAP) need to be standardised although we will
> probably need to discuss the internal PDP algorithms in this working group
> to some extent in order to make sure that the wire-protocols are adequate
> for the job (this is the proof-of-concept stuff that is talked about in
the
> Policy WG charter).
>
I agree that the PDP translation must be left to the implementor, so the
wire protocols ideally should meet the needs of all the scenarios above.  We
really won't know the optimal model until real world deployments begin to
see larger volumes.


Steve
----------------------------------------------------------------
Steve Jackowski
President/CEO
Deterministic Networks, Inc.
245M Mt. Hermon Rd. #140
Scotts Valley, Ca 95066

Office: (831) 421-0388 x18
Home office: (831) 426-4001
Mobile: (408) 813-6294
Fax: (831) 421-0394

> > -----Original Message-----
> > From: Pradosh Kumar Mohapatra [mailto:mohap@research.bell-labs.com]
> > Sent: Friday, July 16, 1999 2:08 PM
> > To: policy
> > Subject: COPS message content
> >
> >
> > Hi,
> > Being new to the mailing-list, I really don't know whether
> > there is any
> > information available on the message content of a COPS
> > message sent from
> > PEP to PDP or PDP to PEP. I am referring to the ClientSI part of the
> > COPS messages. Can't there be any standard way to represent
> > this part of
> > the protocol, given the fact that we have the ldap structure of policy
> > in place.
> > From whatever I have read, I came to understand that this field of the
> > protocol is very much specific to the discipline that it
> > related to e.g.
> > qos, or security. If it is so, then after the packet reaches the PDP,
> > how does the PDP operate? What are the details about how it
> > accesses the
> > LDAP and matches a rule? If a rule is matched, what contents
> > does it put
> > in the reply message?
> >
> > Regards,
> > Pradosh
> >



From majordomo@raleigh.ibm.com  Tue Jul 20 11:33:42 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA06280
	for <policy-archive@odin.ietf.org>; Tue, 20 Jul 1999 11:33:41 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id LAA24878;
	Tue, 20 Jul 1999 11:33:30 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id LAA35374;
	Tue, 20 Jul 1999 11:32:37 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA46174; Tue, 20 Jul 1999 11:14:26 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA46150; Tue, 20 Jul 1999 11:14:22 -0400
Received: from ellesson.raleigh.ibm.com (ellesson.raleigh.ibm.com [9.37.84.214])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id LAA29004
	for <policy@raleigh.ibm.com>; Tue, 20 Jul 1999 11:14:23 -0400
Message-Id: <199907201514.LAA29004@rtpmail01.raleigh.ibm.com>
X-Sender: ellesson@dnjake2.raleigh.ibm.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0
Date: Tue, 20 Jul 1999 11:14:27 -0400
To: policy@raleigh.ibm.com
From: Ed Ellesson <ellesson@raleigh.ibm.com>
Subject: Long:  Ryan Moat's Promised ABNF for Core Schema Doc
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Ed Ellesson <ellesson@raleigh.ibm.com>

This is from Ryan Moats.  I am posting this for him until we work out the
address issue that caused his posting to bounce.  

**************************************************


From: "Ryan Moats" <jayhawk@att.com>
To: "Policy Mailing List" <policy@raleigh.ibm.com>
Subject: LONG (>1K lines): Promised ABNF for Core Schema document...
Date: Tue, 20 Jul 1999 07:12:04 -0500
Message-Id:
<007401bed2a9$14462580$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Importance: Normal


First, my apologies to anybody who sees this twice.
I haven't seen it in the policy archives, so I'm
not sure it got distributed (I hope this isn't a
mail list problem).

Also, I'm on vacation until 7/26, so I won't be
replying to comments until then...

Policy fans-

Here is the ABNF text changes for draft-ietf-policy-core-schema-04
that I promised during the working group meeting in Oslo.

I've got some issues and explanation that I'll summarize here

- all constructs <oid-??#> imply an OID to be assigned. The ?? indicates
  family (for separate arcs) and # is the number.
- In change 2 cn is defined in X.500, rather than define here, it
  needs a reference to RFC 2256.
- The reason changes 3 on looks so complex is that the "AUXILIARY"
  construct is captured in LDAP and X.500 through the use of content
  rules.  Further, the "POSSIBLE SUPERIORS" construct is captured
  through the use of name form rules and structural rules.  For
  each attribute that can function as an RDN, you need a separate
  name form and structural rule.  Finally, the containerStructuralRule
  needs to point at object class container, which is a MicroSoft definition
  that I don't have the OID of.  Do we really want this?
- In change 4, there is no way to define default values for LDAP attributes
  in the schema ABNF.  The default information specified is only
  useful for a management tool and should be moved to a separate section
  or discussion paragraph.
- Using Formatted strings to implement ordering for things with DNs
  policyRuleConditionList, policyRuleActionList, ... is not the right
  approach.  It requires the client to process all strings in the
  response before it can make another LDAP request based on the
  embedded DN as well as a discussion of what the client is supposed
  to do if a returned string is not formatted correctly.  A better
  implementation is to use a separate object class to replace this
  attribute, with the formatted portions of the string being
  attributes.  These separate objects are grouped under the policyRule
  object via DIT containment rules.  The rules supplied take this
  alternate approach.
- Further, in changes 5,6 you cannot specify a possible superior for
  an auxiliary class.  Therefore, that specification has been
  ignored.

It is awfully long, but ABNF can get pretty verbose.

Ryan

============changes

1. replace

     NAME             policy
     DESCRIPTION      An abstract class with four attributes for
                      describing a policy-related instance.
     DERIVED FROM     top
     TYPE             abstract
     AUXILIARY CLASSES    none
     OID              <to be assigned>
     MAY              cn caption description policyKeywords

with

     ( <oid-oc1> NAME 'policy'
       DESC 'An abstract class with four attributes for describing
	     a policy-related instance.'
       SUP top
       ABSTRACT
       MAY (cn $ caption $ description $ policyKeywords)
     )

2. replace

     NAME             caption
     DESCRIPTION      A one-line description of this policy-related
                      object.
     SYNTAX           IA5String

     OID              <to be assigned>
     EQUALITY         caseExactIA5Match
     SINGLE-VALUED

     NAME             description
     DESCRIPTION      A long description of this policy-related object.
     SYNTAX           IA5String
     OID              <to be assigned>
     EQUALITY         caseExactIA5Match
     SINGLE-VALUED

     NAME             policyKeywords
     DESCRIPTION      A set of keywords to assist directory clients in
                      locating the policy objects applicable to them.
                      Standard keyword values are listed in the Policy
                      Core Information Model document.
     SYNTAX           IA5String
     OID              <to be assigned>
     EQUALITY         caseExactIA5Match
     MULTI-VALUED

with
     ( <oid-at1> NAME 'caption'
       DESC 'A one-line description of this policy-related object.'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

     ( <oid-at2> NAME 'description'
       DESC 'A long description of this policy-related object.'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

     ( <oid-at3> NAME 'policyKeywords'
       DESC 'A set of keywords to assist directory clients in locating
	     the policy objects applicable to them. Standard keyword
	     values are listed in the Policy Core Information Model
	     document.'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
     )

3. replace

     NAME             policyGroup
     DESCRIPTION      A container for either a set of related policyRules
                      or a set of related policyGroups.
     DERIVED FROM     top
     TYPE             structural
     AUXILIARY CLASSES    policyGroupContainmentAuxClass,
                      policyRuleContainmentAuxClass
     POSSIBLE SUPERIORS   container, organization, organizationalUnit,
                      policyGroup
     OID              <to be assigned>
     MUST             policyGroupName

   The one attribute of PolicyGroup is defined as:

     NAME             policyGroupName
     DESCRIPTION      The user-friendly name of this policy group.
     SYNTAX           IA5String
     OID              <to be assigned>
     EQUALITY         caseExactIA5Match
     SINGLE-VALUED

with

     ( <oid-oc2> NAME 'policyGroup'
       DESC 'A container for either a set of related policyRules or a
             set of related policyGroups.'
       SUP top
       MUST (policyGroupName)
     )

     ( <oid-cr1>
       NAME 'policyGroupContentRule'
       DESC 'shows what auxiliary classes go with policyGroup'
       AUX (policyGroupContainmentAuxClass $ policyRuleContainmentAuxClass)
     )

     ( <oid-nf1> NAME 'policyGroupNameForm'
       OC policyGroup
       MUST (cn)
     )

     ( 1 NAME 'policyGroupStructuralRule'
       FORM policyGroupNameForm
       SUP 2 3 4 1
     )

     ( <oid-nf2> NAME 'policyGroupNameForm'
       OC policyGroup
       MUST (policyGroupName)
     )


     ( 5 NAME 'policyGroupStructuralRule2'
       FORM policyGroupNameForm
       SUP 2 3 4 5
     )


     ( 2 NAME 'containerStructuralRule'
       FORM ?container object class is MicroSoft specific?
     )

     ( 3 NAME 'organizationStructuralRule'
       FORM 2.5.15.3
     )

     ( 4 NAME 'organizationalUnitStructuralRule'
       FORM 2.5.15.4
     )

   The one attribute of PolicyGroup is defined as:

     ( <oid-at4> NAME 'policyGroupName'
       DESC 'The user-friendly name of this policy group.'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

4. Replace

     NAME             policyRule
     DESCRIPTION      The central class for representing the "If
                      Condition then Action" semantics associated with a
                      policy rule.
     DERIVED FROM     top
     TYPE             structural
     AUXILIARY CLASSES    none
     POSSIBLE SUPERIORS   policyGroup
     OID              <to be assigned>
     MUST             policyRuleName
     MAY              policyRuleEnabled policyRuleConditionListType
                      policyRuleConditionList policyRuleActionList
                      policyRuleValidityPeriodList policyRuleUsage
                      policyRulePriority policyRuleMandatory
                      policyRuleSequencedActions

   The attributes of policyRule are as follows:

     NAME             policyRuleName
     DESCRIPTION      The user-friendly name of this policy rule.
     SYNTAX           IA5String
     OID              <to be assigned>
     EQUALITY         caseExactIA5Match
     SINGLE-VALUED
     NAME             policyRuleEnabled
     DESCRIPTION      An enumeration indicating whether a policy rule is
                      administratively enabled, administratively
                      disabled, or enabled for debug mode.  The defined
                      values for this attribute are enabled(1),
                      disabled(2), and enabledForDebug(3).
     SYNTAX           INTEGER
     OID              <to be assigned>
     EQUALITY         integerMatch
     SINGLE-VALUED
     DEFAULT VALUE    enabled(1)

     NAME             policyRuleConditionListType
     DESCRIPTION      Indicates whether the list of policy conditions
                      associated with this policy rule is in disjunctive
                      normal form (DNF) or conjunctive normal form (CNF).
                      Defined values are DNF(1) and CNF(2).
     SYNTAX           INTEGER
     OID              <to be assigned>
     EQUALITY         integerMatch
     SINGLE-VALUED
     DEFAULT VALUE    DNF(1)

     NAME             policyRuleConditionList
     DESCRIPTION      An unordered list of strings of the form
                      '<groupNumber>:<+|->:<DN>', indicating a set of
                      policy conditions that determine when the
                      policyRule is applicable.
     SYNTAX           DirectoryString
     OID              <to be assigned>
     EQUALITY         caseIgnoreSubstringsMatch
     MULTI-VALUED
     FORMAT           <groupNumber>:<+|->:<DN>

     NAME             policyRuleActionList
     DESCRIPTION      An unordered list of strings of the form 'n:DN',
                      indicating an ordered set of policy actions to be

                      performed if the associated condition(s) of the
                      policyRule evaluates to true.
     SYNTAX           DirectoryString
     OID              <to be assigned>
     EQUALITY         caseIgnoreSubstringsMatch
     MULTI-VALUED
     FORMAT           n:DN

     NAME             policyRuleValidityPeriodList
     DESCRIPTION      Distinguished names of policyTimePeriodConditions
                      that determine when the policyRule is scheduled to
                      be active / inactive.  No order is implied.
     SYNTAX           DN
     OID              <to be assigned>
     EQUALITY         distinguishedNameMatch
     MULTI-VALUED


     NAME             policyRuleUsage
     DESCRIPTION      This attribute is used to provide guidelines on how
                      this policy should be used.
     SYNTAX           DirectoryString
     OID              <to be assigned>
     EQUALITY         caseIgnoreMatch
     SINGLE-VALUED


     NAME             policyRulePriority
     DESCRIPTION      A non-negative integer for prioritizing this
                      policyRule relative to other policyRules.  A larger
                      value indicates a higher priority.
     SYNTAX           INTEGER
     OID              <to be assigned>
     EQUALITY         integerMatch
     SINGLE-VALUED
     DEFAULT VALUE    0


     NAME             policyRuleMandatory
     DESCRIPTION      A flag indicating that the evaluation of the
                      policyConditions and execution of policyActions (if
                      the condition list evaluates to True) is required.
     SYNTAX           Boolean
     OID              <to be assigned>
     EQUALITY         booleanMatch
     SINGLE-VALUED
     DEFAULT VALUE    TRUE


     NAME             policyRuleSequencedActions
     DESCRIPTION      An enumeration indicating how to interpret the
                      action ordering indicated via the
                      policyRuleActionList attribute.  The defined values
                      for this attribute are mandatory(1),
                      recommended(2), and dontCare(3).

     SYNTAX           INTEGER
     OID              <to be assigned>
     EQUALITY         integerMatch
     SINGLE-VALUED
     DEFAULT VALUE    dontCare(3)

with
     ( <oid-oc3> NAME 'policyRule'
       DESC 'The central class for representing the "If Condition then
Action" semantics associated with a policy rule.'
       SUP top
       MUST (policyRuleName)
       MAY (policyRuleEnabled $ policyRuleConditionListType $
	    policyRuleValidityPeriodList $ policyRuleUsage $
            policyRulePriority $ policyRuleMandatory $
            policyRuleSequencedActions)
     )

     ( <oid-nf3> NAME 'policyRuleNameForm'
       OC policyRule
       MUST (cn)
     )

     ( 6 NAME 'policyRuleStructuralRule'
         FORM policyRuleNameForm
         SUP 5
     )

     ( <oid-nf4> NAME 'policyRuleNameForm'
       OC policyRule
       MUST (policyRuleName)
     )

     ( 7 NAME 'policyRuleStructuralRule2'
         FORM policyRuleNameForm
         SUP 5

     )

     ( <oid-at5> NAME 'policyRuleName'
       DESC 'The user-friendly name of this policy rule.'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

     ( <oid-at6> NAME 'policyRuleEnabled'
       DESC 'An enumeration indicating whether a policy rule is
             administratively enabled, administratively disabled, or enabled
             for debug mode. The defined values for this attribute are
             enabled(1), disabled(2), and enabledForDebug(3).'
       SYNTAX INTEGER
       EQUALITY integerMatch
       SINGLE-VALUE
     )

     ( <oid-at7> NAME 'policyRuleConditionListType'
       DESC 'Indicates whether the list of policy conditions
             associated with this policy rule is in disjunctive normal
             form (DNF) or conjunctive normal form (CNF). Defined
	     values are DNF(1) and CNF(2).'
       SYNTAX INTEGER
       EQUALITY integerMatch
       SINGLE-VALUE
     )

     ( <oid-at8> NAME 'policyRuleValidityPeriodList'
       DESC 'An unordered list of strings of the form "n:DN",
             indicating an ordered set of policy actions to be performed if
	     the associated condition(s) of the policyRule evaluates to
	     true.'
       SYNTAX DirectoryString
       EQUALITY caseIgnoreSubstringsMatch
     )

     ( <oid-at9> NAME 'policyRuleUsage'
       DESC 'This attribute is used to provide guidelines on how this
             policy should be used.'
       SYNTAX DirectoryString
       EQUALITY caseIgnoreMatch
       SINGLE-VALUE
     )

     ( <oid-at10> NAME 'policyRulePriority'
       DESC 'A non-negative integer for prioritizing this policyRule
             relative to other policyRules. A larger value indicates a
	     higher priority.'
       SYNTAX INTEGER
       EQUALITY integerMatch
       SINGLE-VALUE
     )

     ( <oid-at11> NAME 'policyRuleMandatory'
       DESC 'A flag indicating that the evaluation of the
             policyConditions and execution of policyActions (if the
	     condition list evaluates to True) is required.'
       SYNTAX Boolean
       EQUALITY booleanMatch
       SINGLE-VALUE
     )

     ( <oid-at12> NAME 'policyRuleSequencedActions'
       DESC 'An enumeration indicating how to interpret the action
             ordering indicated via the policyRuleActionList
             attribute. The defined values for this attribute are
             mandatory(1), recommended(2), and dontCare(3).'
       SYNTAX Boolean
       EQUALITY booleanMatch
     )

     ( <oid-oc4> NAME 'policyRuleConditionItem'
       DESC 'The class contains attributes indicating a set of policy
             conditions that determine when the policyRule is
             applicable.'
       SUP top
       MUST (policyRuleItemGroupNumber $ policyRuleItemNegated $
             policyRuleItemDN $ policyRuleConditionItemName)
     )

     ( <oid-nf5> NAME 'policyRuleConditionItemNameForm'
       OC policyRuleConditionItem
       MUST (cn)
     )

     ( 8 NAME 'policyRuleConditionItemStructuralRule'
       FORM policyRuleConditionItemNameForm
       SUP 7
     )


     ( <oid-nf6> NAME 'policyRuleConditionItemNameForm'
       OC policyRuleConditionItem
       MUST (policyRuleConditionItemName)
     )

     ( 9 NAME 'policyRuleConditionItemStructuralRule2'
       FORM policyRuleConditionItemNameForm
       SUP 7
     )

     ( <oid-at13>
       NAME 'policyRuleConditionItemName'
       DESC 'a user-friendly name for the policyRuleItem'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

     ( <oid-at14>
       NAME 'policyRuleItemGroupNumber'
       DESC 'The group number a policyRuleConditionItem or
             policyRuleActionItem is a member of'
       SYNTAX INTEGER
       EQUALITY integerMatch
       SINGLE-VALUE
     )

     ( <oid-at15>
       NAME 'policyRuleItemNegated'
       DESC 'Indicates if a policyRuleConditionItem is negated TRUE =
             negation'
       SYNTAX BOOLEAN
       EQUALITY integerMatch
       SINGLE-VALUE
     )

     ( <oid-at16>
       NAME 'policyRuleItemDN'
       DESC 'a DN pointer to the policy condition or policy action'
       SYNTAX DN
       EQUALITY distinguishedNameMatch
       SINGLE-VALUE
     )

     ( <oid-oc5> NAME 'policyRuleActionItem'
       DESC 'The class contains attributes indicating a set of policy
             actions to be performed if the associated condition(s) of the
             policy Rule evaluates to true.'
       SUP top
       MUST (policyRuleItemGroupNumber $ policyRuleItemDN $
             policyRuleActionItemName)
     )

     ( <oid-nf7> NAME 'policyRuleActionItemNameForm'
       OC policyRuleActionItem
       MUST (cn)
     )

     ( 10 NAME 'policyRuleActionItemStructuralRule'
       FORM policyRuleActionItemNameForm
       SUP 7
     )

     ( <oid-nf8> NAME 'policyRuleActionItemNameForm'
       OC policyRuleActionItem
       MUST (policyRuleActionItemName)
     )

     ( 11 NAME 'policyRuleActionItemStructuralRule2'
       FORM policyRuleActionItemNameForm
       SUP 7
     )

     ( <oid-at17>
       NAME 'policyRuleActionItemName'
       DESC 'a user-friendly name for the policyRuleItem'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

5. Replace

     NAME             policyCondition
     DESCRIPTION      A class representing a condition to be evaluated in
                      conjunction with a policy rule.
     DERIVED FROM     top
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     POSSIBLE SUPERIORS   policyRule
     OID              <to be assigned>
     MUST             policyConditionName
     MAY


   The one attribute definition is as follows:

     NAME             policyConditionName
     DESCRIPTION      The user-friendly name of this policy condition.
     SYNTAX           IA5String
     OID              <to be assigned>
     EQUALITY         caseExactIA5Match
     SINGLE-VALUED

with

     ( <oid-oc6> NAME 'policyCondition'
       DESC 'A class representing a condition to be evaluated in
             conjunction with a policy rule.'
       SUP top
       AUXILIARY
       MUST (policyConditionName)
     )


     ( <oid-at18>
       NAME 'policyConditionName'
       DESC 'The user-friendly name of this policy condition.'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

6. Replace

     NAME             policyTimePeriodCondition
     DESCRIPTION      A class that provides the capability of enabling /
                      disabling a policy rule according to a pre-
                      determined schedule.
     DERIVED FROM     policyCondition
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     POSSIBLE SUPERIORS   policyRule
     OID              <to be assigned>
     MUST
     MAY              ptpConditionTime ptpConditionMonthOfYearMask
                      ptpConditionDayOfMonthMask
                      ptpConditionDayOfWeekMask ptpConditionTimeOfDayMask
                      ptpConditionTimeZone

   The attributes of policyTimePeriodCondition are defined as follows:

     NAME             ptpConditionTime
     DESCRIPTION      The range of calendar dates on which a policy rule
                      is valid.
     SYNTAX           PrintableString
     OID              <to be assigned>
     EQUALITY         caseIgnoreMatch
     SINGLE-VALUED
     FORMAT           [yyyymmddhhmmss]:[yyyymmddhhmmss]


     NAME             ptpConditionMonthOfYearMask
     DESCRIPTION      A mask identifying the months of the year in which
                      a policy rule is valid.
     SYNTAX           Printable String
     OID              <to be assigned>
     EQUALITY         caseIgnoreMatch
     SINGLE-VALUED
     FORMAT           A string of 12 ASCII '0's and '1's.


     NAME             ptpConditionDayOfMonthMask
     DESCRIPTION      A mask identifying the days of the month on which a
                      policy rule is valid.
     SYNTAX           PrintableString
     OID              <to be assigned>
     EQUALITY         caseIgnoreMatch
     SINGLE-VALUED
     FORMAT           A string of 31 ASCII '0's and '1's.


     NAME             ptpConditionDayOfWeekMask
     DESCRIPTION      A mask identifying the days of the week on which a
                      policy rule is valid.
     SYNTAX           PrintableString
     OID              <to be assigned>
     EQUALITY         caseIgnoreMatch
     SINGLE-VALUED
     FORMAT           A string of 7 ASCII '0's and '1's.


     NAME             ptpConditionTimeOfDayMask
     DESCRIPTION      The range of times at which a policy rule is valid.
                      If the second time is earlier than the first, then
                      the interval spans midnight.
     SYNTAX           Printable String
     OID              <to be assigned>
     EQUALITY         caseIgnoreMatch
     SINGLE-VALUED
     FORMAT           hhmmss:hhmmss

     NAME             ptpConditionTimeZone
     DESCRIPTION      The definition of the time zone for the
                      policyTimePeriodCondition.
     SYNTAX           PrintableString
     OID              <to be assigned>
     EQUALITY         caseIgnoreMatch
     SINGLE-VALUED
     FORMAT           either 'Z' (UTC) or <'+'|'-'><hhmm>


with

     ( <oid-oc7> NAME 'policyTimePeriodCondition'
       DESC 'A class that provides the capability of enabling /
             disabling a policy rule according to a pre- determined
	     schedule.'
       SUP policyCondition
       AUXILIARY
       MAY (ptpConditionTime $ ptpConditionMonthOfYearMask $
            ptpConditionDayOfMonthMask $ ptpConditionDayOfWeekMask $
	    ptpConditionTimeOfDayMask $ ptpConditionTimeZone)
     )

     ( <oid-at19>
       NAME 'ptpConditionTime'
       DESC 'The range of calendar dates on which a policy rule is valid.'
       SYNTAX PrintableString
       EQUALITY caseIgnoreMatch
       SINGLE-VALUE
     )

     ( <oid-at20>
       NAME 'ptpConditionMonthOfYearMask'
       DESC 'A mask identifying the months of the year in which a
             policy rule is valid.'
       SYNTAX Printable
       EQUALITY caseIgnoreMatch
       SINGLE-VALUE
     )

     ( <oid-at21>
       NAME 'ptpConditionDayOfMonthMask'
       DESC 'A mask identifying the days of the month on which a
             policy rule is valid.'
       SYNTAX PrintableString
       EQUALITY caseIgnoreMatch
       SINGLE-VALUE
     )

     ( <oid-at22>
       NAME 'ptpConditionDayOfWeekMask'
       DESC 'A mask identifying the days of the week on which a policy
             rule is valid.'
       SYNTAX PrintableString
       EQUALITY caseIgnoreMatch
       SINGLE-VALUE
     )

     ( <oid-at23>
       NAME 'ptpConditionTimeOfDayMask'
       DESC 'The range of times at which a policy rule is valid. If
             the second time is earlier than the first, then the interval
             spans midnight.'
       SYNTAX Printable
       EQUALITY caseIgnoreMatch
     )

     ( <oid-at24>
       NAME 'ptpConditionTimeZone'
       DESC 'The definition of the time zone for the
             policyTimePeriodCondition.'
       SYNTAX PrintableString
       EQUALITY caseIgnoreMatch
       SINGLE-VALUE
     )

7. Replace

     NAME             vendorPolicyCondition
     DESCRIPTION      A class that defines a registered means to describe
                      a policy condition.
     DERIVED FROM     policyCondition
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     POSSIBLE SUPERIORS   policyRule
     OID              <to be assigned>
     MUST             vendorPolicyConstraintData
                      vendorPolicyConstraintEncoding
     MAY

   The attribute definitions for vendorPolicyCondition are as follows:

     NAME             vendorPolicyConstraintData
     DESCRIPTION      Escape mechanism for representing constraints that
                      have not been modeled as specific attributes. The
                      format of the values is identified by the OID
                      stored in the attribute
                      vendorPolicyConstraintEncoding.
     SYNTAX           OctetString
     OID              <to be assigned>
     EQUALITY         octetStringMatch
     MULTI-VALUED


     NAME             vendorPolicyConstraintEncoding
     DESCRIPTION      An OID identifying the format and semantics for
                      this instance's vendorPolicyConstraintData

                      attribute.
     SYNTAX           OID
     OID              <to be assigned>
     EQUALITY         objectIdentifierMatch
     SINGLE-VALUED

with

     ( <oid-oc8> NAME 'vendorPolicyCondition'
       DESC 'A class that defines a registered means to describe a
	     policy condition.'
       SUP policyCondition
       AUXILIARY
       MUST (vendorPolicyConstraintData $
	     vendorPolicyConstraintEncoding)
     )

     ( <oid-at25>
       NAME 'vendorPolicyConstraintData'
       DESC 'Escape mechanism for representing constraints that have
	    not been modeled as specific attributes. The format of the
	    values is identified by the OID stored in the attribute
	    vendorPolicyConstraintEncoding.'
       SYNTAX OctetString
       EQUALITY octetStringMatch
     )

     ( <oid-at26>
       NAME 'vendorPolicyConstraintEncoding'
       DESC 'An OID identifying the format and semantics for this
             instance"s vendorPolicyConstraintData attribute.'
       SYNTAX OID
       EQUALITY objectIdentifierMatch
       SINGLE-VALUE
     )

8. Replace

     NAME             policyAction
     DESCRIPTION      A class representing an action to be performed as a
                      result of a policy rule.
     DERIVED FROM     top
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     POSSIBLE SUPERIORS   policyRule
     OID              <to be assigned>
     MUST             policyActionName
     MAY


   The one attribute definition is as follows:

     NAME             policyActionName
     DESCRIPTION      The user-friendly name of this policy action.
     SYNTAX           IA5String
     OID              <to be assigned>
     EQUALITY         caseExactIA5Match
     SINGLE-VALUED

with

     ( <oid-oc9> NAME 'policyAction'
       DESC 'A class representing an action to be performed as a
             result of a policy rule.'
       SUP top
       AUXILIARY
       MUST (policyActionName)
     )

     ( <oid-at27>
       NAME 'policyActionName'
       DESC 'The user-friendly name of this policy action.'
       SYNTAX IA5String
       EQUALITY caseExactIA5Match
       SINGLE-VALUE
     )

9. Replace

     NAME             vendorPolicyAction
     DESCRIPTION      A class that defines a registered means to describe
                      a policy action.
     DERIVED FROM     policyAction
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     POSSIBLE SUPERIORS   policyRule
     OID              <to be assigned>
     MUST             vendorPolicyActionData vendorPolicyActionEncoding
     MAY

   The attribute definitions for vendorPolicyAction are as follows:

     NAME             vendorPolicyActionData
     DESCRIPTION      Escape mechanism for representing actions that have
                      not been modeled as specific attributes. The
                      format of the values is identified by the OID
                      stored in the attribute vendorPolicyActionEncoding.
     SYNTAX           OctetString
     OID              <to be assigned>
     EQUALITY         octetStringMatch
     MULTI-VALUED



     NAME             vendorPolicyActionEncoding
     DESCRIPTION      An OID identifying the format and semantics for
                      this instance's vendorPolicyActionData attribute.
     SYNTAX           OID
     OID              <to be assigned>
     EQUALITY         objectIdentifierMatch
     SINGLE-VALUED

with

     ( <oid-oc10> NAME 'vendorPolicyAction'
       DESC 'A class that defines a registered means to describe a
             policy action.'
       SUP policyAction
       AUXILIARY
       MUST (vendorPolicyActionData $ vendorPolicyActionEncoding)
     )

     ( <oid-at28>
       NAME 'vendorPolicyActionData'
       DESC 'Escape mechanism for representing actions that have not
	     been modeled as specific attributes. The format of the
	     values is identified by the OID stored in the attribute
	     vendorPolicyActionEncoding.'
       SYNTAX OctetString
       EQUALITY octetStringMatch
     )

     ( <oid-at29>
       NAME 'vendorPolicyActionEncoding'
       DESC 'An OID identifying the format and semantics for this
             instance"s vendorPolicyActionData attribute.'
       SYNTAX OID
       EQUALITY objectIdentifierMatch
       SINGLE-VALUE
     )

10. Replace

     NAME           policyInstance
     DESCRIPTION    A structural class to which the auxiliary classes
                    policyCondition and policyAction may be attached,
                    when the complex policy rule structure is required.
     DERIVED FROM   policy
     TYPE           structural
     AUXILIARY CLASSES    policyCondition, policyAction
     OID            <to be assigned>

with

     ( <oid-oc11> NAME 'policyInstance'
       DESC 'A structural class to which the auxiliary classes
	     policyCondition and policyAction may be attached, when
	     the complex policy rule structure is required.'
       SUP policy
     )

     ( <oid-cr2>
       NAME 'policyInstanceContentRule'
       DESC 'shows what auxiliary classes go with policyInstance'
       AUX (policyCondition $ policyAction)
     )

11. Replace

     NAME             policyElement
     DESCRIPTION      An auxiliary class used to tag instances of classes
                      defined outside the realm of policy as relevant to
                      a particular policy specification.
     DERIVED FROM     policy
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     OID              <to be assigned>

with

     ( <oid-oc12> NAME 'policyElement'
       DESC 'An auxiliary class used to tag instances of classes
	     defined outside the realm of policy as relevant to a
	     particular policy specification.'
       SUP policy
       AUXILIARY
     )

12. Replace

     NAME             policySubtreesPtrAuxClass
     DESCRIPTION      An auxiliary class providing a DN pointer to roots
                      of DIT subtrees contained policy-related objects.
     DERIVED FROM     top
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     OID              <to be assigned>
     MUST             policySubtreesAuxContainedSet

with

     ( <oid-oc13> NAME 'policySubtreesPtrAuxClass'
       DESC 'An auxiliary class providing a DN pointer to roots of DIT

             subtrees contained policy-related objects.'
       SUP top
       AUXILIARY
       MUST (policySubtreesAuxContainedSet)
     )

13. Replace

     NAME             policySubtreesAuxContainedSet
     DESCRIPTION      Distinguished names of objects that serve as roots
                      for DIT subtrees containing policy-related objects.
                      No order is implied.
     SYNTAX           DN
     OID              <to be assigned>
     EQUALITY         distinguishedNameMatch
     MULTI-VALUED

with

     ( <oid-at30>
       NAME 'policySubtreesAuxContainedSet'
       DESC 'Distinguished names of objects that serve as roots for
             DIT subtrees containing policy-related objects. No order
	     is implied.'
       SYNTAX DN
       EQUALITY distinguishedNameMatch
     )

14. Replace

     NAME                 policyGroupContainmentAuxClass
     DESCRIPTION          An auxiliary class used to bind policyGroups
                          to an appropriate container object.
     DERIVED FROM         top
     TYPE                 auxiliary
     AUXILIARY CLASSES    none
     OID                  <to be assigned>
     MUST                 policyGroupsAuxContainedSet

with

     ( <oid-oc14> NAME 'policyGroupContainmentAuxClass'
       DESC 'An auxiliary class used to bind policyGroups to an
             appropriate container object.'
       SUP top
       AUXILIARY
       MUST (policyGroupsAuxContainedSet)
     )

15. Replace

     NAME           policyGroupsAuxContainedSet
     DESCRIPTION    Distinguished names of policyGroups associated in
                    some way with the instance to which this attribute
                    has been appended.  No order is implied.
     SYNTAX         DN
     OID            <to be assigned>
     EQUALITY       distinguishedNameMatch
     MULTI-VALUED

with

     ( <oid-at31>
       NAME 'policyGroupsAuxContainedSet'
       DESC 'Distinguished names of policyGroups associated in some
             way with the instance to which this attribute has been
             appended. No order is implied.'
       SYNTAX DN
       EQUALITY distinguishedNameMatch
     )

16.  Replace

     NAME             policyRuleContainmentAuxClass
     DESCRIPTION      An auxiliary class used to bind policyRules to an
                      appropriate container object.
     DERIVED FROM     top
     TYPE             auxiliary
     AUXILIARY CLASSES    none
     OID              <to be assigned>
     MUST             policyRulesAuxContainedSet

with

     ( <oid-oc15> NAME 'policyRuleContainmentAuxClass'
       DESC 'An auxiliary class used to bind policyRules to an
             appropriate container object.'
       SUP top
       AUXILIARY
       MUST (policyRulesAuxContainedSet)
     )

17.  Replace

     NAME             policyRulesAuxContainedSet
     DESCRIPTION      Distinguished names of policyRules associated in
                      some way with the instance to which this attribute
                      has been appended.  No order is implied.
     SYNTAX           DN
     OID              <to be assigned>
     EQUALITY         distinguishedNameMatch

     MULTI-VALUED

with

     ( <oid-at32>
       NAME 'policyRulesAuxContainedSet'
       DESC 'Distinguished names of policyRules associated in some way
             with the instance to which this attribute has been
             appended. No order is implied.'
       SYNTAX DN
       EQUALITY distinguishedNameMatch
     )

Regards, 

Ed Ellesson
Sr. Engineer, TCP/IP Technology Management
IBM Networking Software Products
Research Triangle Park, NC 
ellesson@raleigh.ibm.com
919-254-4115


From majordomo@raleigh.ibm.com  Tue Jul 20 12:27:08 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA07367
	for <policy-archive@odin.ietf.org>; Tue, 20 Jul 1999 12:27:05 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA38104;
	Tue, 20 Jul 1999 12:26:47 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id MAA19902;
	Tue, 20 Jul 1999 12:25:15 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA32968; Tue, 20 Jul 1999 12:09:13 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA51904; Tue, 20 Jul 1999 12:09:11 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id MAA32030
	for <policy@raleigh.ibm.com>; Tue, 20 Jul 1999 12:09:13 -0400
Received: from sol.extremenetworks.com (sol.extremenetworks.com [216.52.8.2])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA17772
	for <policy@raleigh.ibm.com>; Tue, 20 Jul 1999 12:09:07 -0400
Received: by ftp.extremenetworks.com with Internet Mail Service (5.5.2448.0)
	id <36G72TQH>; Tue, 20 Jul 1999 09:08:39 -0700
Message-Id: <D0805D3B448BD211A7990008C7B18130143078@ftp.extremenetworks.com>
From: Andrew Smith <andrew@extremenetworks.com>
To: "'stevej@DeterministicNetworks.com '"
	 <stevej@DeterministicNetworks.com>
Cc: "'rap (E-mail) '" <rap@iphighway.com>,
        "'policy '"
	 <policy@raleigh.ibm.com>
Subject: RE: COPS message content
Date: Tue, 20 Jul 1999 09:08:39 -0700
Mime-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Andrew Smith <andrew@extremenetworks.com>

Steve,

You wrote:
> Today there are vendors who think these are three
> separate entities, others who think the Policy Server and PDP are 
> combined, and still others who think the PDP and PEP should be combined.
> In my personal opinion, the first has scaling problems with per-flow 
> requests to the PDP.  The second has scaling problems with PEP requests 
> to the Policy Server, and only the third seems uniformly scalable - 
> policies and associated rules are distributed to remote PDP/PEP entities
> from the repository.  

I think I might take issue with your statement that the combined PDP/PEP is
the most scalable solution if you are talking about using LDAP or SNMP to
distribute the rules - but it is early days now and there are not that many
implementations around yet, particularly of LDAP, on which to base opinions.
But I just wanted to point out that I think there's a lot we can do,
particularly with policy-controlled COPS/RSVP admission, in order to improve
the scalability.

I tried to present some ideas on this in ISSLL and RAP WGs at the March IETF
in Minneapolis (draft at
http://www.ietf.org/internet-drafts/draft-smith-sbm-config-00.txt and slides
at http://www.ietf.org/proceedings/99mar/slides/rap-cops-99mar/index.html)
but I don't think I got my point across very well: primarily that there are
likely to be many simple policy control decisions that can be performed
easily by a router/switch/host and do not need to be "out-sourced" and there
will be other decisions that do. i.e. a co-located PDP/PEP can handle simple
decisions but they will likely need to be split for more complex ones. You
then need some sort of "meta-rules" to decide which decisions can be local
and which not: these meta-rules were the basis for most of the draft (I took
this all out when I turned it into just an SBM MIB
draft-ietf-issll-is802-sbm-mib-00.txt). I'm just wondering how to resurrect
this for future work.

Andrew


From majordomo@raleigh.ibm.com  Mon Jul 26 13:47:36 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA17589
	for <policy-archive@odin.ietf.org>; Mon, 26 Jul 1999 13:47:35 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA18308;
	Mon, 26 Jul 1999 13:47:24 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id NAA38068;
	Mon, 26 Jul 1999 13:47:10 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA29472; Mon, 26 Jul 1999 13:33:07 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA52730; Mon, 26 Jul 1999 13:33:02 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id NAA28912
	for <policy@raleigh.ibm.com>; Mon, 26 Jul 1999 13:33:04 -0400
Received: from alms1.fw.att.com (alms1.att.com [192.128.167.146])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA31412
	for <policy@raleigh.ibm.com>; Mon, 26 Jul 1999 13:33:01 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by alms1.fw.att.com (AT&T IPNS/MS-2.2) with SMTP id NAA16946
	for <policy@raleigh.ibm.com>; Mon, 26 Jul 1999 13:32:35 -0400 (EDT)
Received: from schooner.local.windrose.omaha.ne.us by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id NAA27004; Mon, 26 Jul 1999 13:32:33 -0400
From: "Ryan Moats" <jayhawk@att.com>
To: <policy@raleigh.ibm.com>
Subject: ABNF and Versioning (or why Versioning isn't needed)
Date: Mon, 26 Jul 1999 12:30:36 -0500
Message-Id: <000401bed78c$92953d80$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
In-Reply-To: <199907201514.LAA29004@rtpmail01.raleigh.ibm.com>
Importance: Normal
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Ryan Moats" <jayhawk@att.com>
Content-Transfer-Encoding: 7bit

At the Oslo IETF, I promised ABNF for the core
schema document.  That's been included in
a previous mail.  Once the QoS schema
is relatively finalized, I'm willing to volunteer
ABNF for that as well.

I also volunteered schema versioning information
in Oslo. If the ABNF that I sent is accepted, then
schema versioning isn't as necessary.  That's because
the ABNF I sent does not use structured DNs.  Rather,
it uses DIT containment rules (which in LDAP are
specified on a one-level basis) to group the rule
and action items under the policy rule.

This still allows clients to pull all policy rule objects in a single LDAP
query (all policy item
objects will contain the RDN of the containing policy rule object) as well
as being in line with work items
in LDAPEXT and X.500 (as well as elsewhere).

Ryan



From majordomo@raleigh.ibm.com  Tue Jul 27 04:52:24 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id EAA18547
	for <policy-archive@odin.ietf.org>; Tue, 27 Jul 1999 04:52:23 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id EAA27100;
	Tue, 27 Jul 1999 04:52:16 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id EAA40182;
	Tue, 27 Jul 1999 04:52:07 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA29652; Tue, 27 Jul 1999 04:41:11 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA28948; Tue, 27 Jul 1999 04:40:52 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id EAA26698
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 04:40:52 -0400
Received: from csi-admin1.cisco.com (csi-admin1.cisco.com [144.254.79.12])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id EAA31498
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 04:40:50 -0400
Received: from yramberghpxu (telaviv2-dhcp11.cisco.com [144.254.93.61]) by csi-admin1.cisco.com (8.8.4-Cisco.1/8.6.5) with SMTP id LAA13239; Tue, 27 Jul 1999 11:42:12 +0300 (IDT)
Message-Id: <008601bed84d$e5bf6640$3d5dfe90@cisco.com>
From: "Yoram Ramberg" <yramberg@cisco.com>
To: "Ryan Moats" <jayhawk@att.com>, <policy@raleigh.ibm.com>
Cc: "Yoram Snir" <ysnir@cisco.com>
References: <000401bed78c$92953d80$e3c8090a@schooner.local.windrose.omaha.ne.us>
Subject: Re: ABNF and Versioning (or why Versioning isn't needed)
Date: Tue, 27 Jul 1999 11:34:26 -0500
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Yoram Ramberg" <yramberg@cisco.com>
Content-Transfer-Encoding: 7bit

Ryan - thanx much! We've been thinking about that and were hoping you'd be
willing to help.
Once the QoS schema is close to completion we'll get in touch. There are
certain modifications being considered to the core schema. The idea is to
follow up on recommendations made in Oslo and move some of the features
presented as part of the QoS schema to the core schema. We are working on
this proposal with Bob, Ed and John who are the authors of the core schema.

I hope that by the time of the next IETF meeting we'll have the revised
proposal (for core and QoS schemata and information models) where all the
schema definition is ABNF. Your help will be invaluable and appreciated.

On the switch from SDN to RDN containment:

This may be tricky. See if I got this one right.
Once you impose strict containment on the DIT you're also presenting a
restriction on relationships between rules and conditions relationships.
What happens is that your forcing me to place all conditions related to a
given rule under this rule in the DIT. What if another rule "wants" to use
one of those conditions? It must copy that condition (if it knows how to
find it) and place the copy unde the new rule. This makes that condition
non-reusable. Two identical conditions must be maintained now and could be
modified independently.

(I used conditions in my example because we deem conditions to be prime
candidate for reusability. Everyone wishing to create a reusable object
class and place it in the repository will have this problem exactly)

Furthermore: SDN provide syntax for specifying the intra-expression
relations, i.e., c1 AND c2 and NOT c3 or (NOT c4 abd C7). How would you
express this if you drop SDNs altogether? We, then, must add syntax for this
kind of expressions. RDN doesn't do the trick but atructured RDN still may
(still doesn't solve reusability).

It's quite probable that I just didn't undestand what you wrote and made a
big fuss out of nothing. Please, explain.

Thanx muxh,
*Yoram


----- Original Message -----
From: Ryan Moats <jayhawk@att.com>
To: <policy@raleigh.ibm.com>
Sent: Monday, July 26, 1999 12:30 PM
Subject: ABNF and Versioning (or why Versioning isn't needed)


> At the Oslo IETF, I promised ABNF for the core
> schema document.  That's been included in
> a previous mail.  Once the QoS schema
> is relatively finalized, I'm willing to volunteer
> ABNF for that as well.
>
> I also volunteered schema versioning information
> in Oslo. If the ABNF that I sent is accepted, then
> schema versioning isn't as necessary.  That's because
> the ABNF I sent does not use structured DNs.  Rather,
> it uses DIT containment rules (which in LDAP are
> specified on a one-level basis) to group the rule
> and action items under the policy rule.
>
> This still allows clients to pull all policy rule objects in a single LDAP
> query (all policy item
> objects will contain the RDN of the containing policy rule object) as well
> as being in line with work items
> in LDAPEXT and X.500 (as well as elsewhere).
>
> Ryan
>
>



From majordomo@raleigh.ibm.com  Tue Jul 27 09:40:29 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA03514
	for <policy-archive@odin.ietf.org>; Tue, 27 Jul 1999 09:40:23 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id JAA22012;
	Tue, 27 Jul 1999 09:39:48 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id JAA38570;
	Tue, 27 Jul 1999 09:38:35 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA55642; Tue, 27 Jul 1999 09:28:02 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA38738; Tue, 27 Jul 1999 09:28:00 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id JAA39110
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 09:28:02 -0400
Received: from alms1.fw.att.com (alms1.att.com [192.128.167.146])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id JAA38490
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 09:28:00 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by alms1.fw.att.com (AT&T IPNS/MS-2.2) with SMTP id JAA13944;
	Tue, 27 Jul 1999 09:27:37 -0400 (EDT)
Received: from schooner.local.windrose.omaha.ne.us by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id JAA16913; Tue, 27 Jul 1999 09:27:35 -0400
From: "Ryan Moats" <jayhawk@att.com>
To: "Yoram Ramberg" <yramberg@cisco.com>, <policy@raleigh.ibm.com>
Cc: "Yoram Snir" <ysnir@cisco.com>
Subject: RE: ABNF and Versioning (or why Versioning isn't needed)
Date: Tue, 27 Jul 1999 08:25:35 -0500
Message-Id: <001501bed833$827fd520$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
In-Reply-To: <008601bed84d$e5bf6640$3d5dfe90@cisco.com>
Importance: Normal
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Ryan Moats" <jayhawk@att.com>
Content-Transfer-Encoding: 7bit

[Some editing to cut down B/W]

| From: Yoram Ramberg [mailto:yramberg@cisco.com]
|
| On the switch from SDN to RDN containment:
|
| This may be tricky. See if I got this one right.
| Once you impose strict containment on the DIT you're also presenting a
| restriction on relationships between rules and conditions relationships.
| What happens is that your forcing me to place all conditions related to a
| given rule under this rule in the DIT. What if another rule "wants" to use
| one of those conditions? It must copy that condition (if it knows how to
| find it) and place the copy unde the new rule. This makes that condition
| non-reusable. Two identical conditions must be maintained now and could be
| modified independently.

Yes and No (and now let me explain).  What the ABNF specifies
(and what is the correct way, I might add) is that
policyRuleConditionItem and policyRuleActionItem (two new
object classes) are restricted to be contained under the
policyRule object class.  For policyRuleConditionItem,
there are four attributes defined: policyRuleItemGroupNumber,
policyRuleItemNegated, policyRuleItemDN, and
policyRuleConditionItemName.  The last is a friendly name that
MAY be used as an RDN.  The other three are separate attributes
that replace the structures of the SDN as separate attributes.
Now, while a tuple (group number, negation, DN) [(group number, DN)
for policyRuleActionItem] is bound under a single policyRule in
the DIT, other policyRules may reuse the DN pointer to build
there own policy items.  While it is possible to restructure the
ABNF to allow for entire policyRuleConditionItem and policyRuleActionItem
reuse, I'm not yet convinced that this is a good thing.  Reusing
the DN's achieves 80-90% (my estimate) of full object class reuse
without cluttering up the schema with extra DN pointers.

| (I used conditions in my example because we deem conditions to be prime
| candidate for reusability. Everyone wishing to create a reusable object
| class and place it in the repository will have this problem exactly)
|
| Furthermore: SDN provide syntax for specifying the intra-expression
| relations, i.e., c1 AND c2 and NOT c3 or (NOT c4 abd C7). How would you
| express this if you drop SDNs altogether? We, then, must add
| syntax for this kind of expressions. RDN doesn't do the trick
| but atructured RDN still may still doesn't solve reusability).

The attribute policyRuleItemNegated handles the negation on an object by
object basis.

SDNs (or sRDNs) are a bad idea because they require the
client to process the returned string BEFORE they can determine if they
need to make another LDAP query to retrieve the pointed to object.  This
means that a client wanting to make only a single LDAP query would be
required to pull the ENTIRE policy tree [this is a BAD thing].
Also, by separating out the DN, you are setting up to take advantage of
a work that is going on in LDAPEXT.  If you use SDNs, you are limited
to "rolling your own".

Ryan




From majordomo@raleigh.ibm.com  Tue Jul 27 11:25:17 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA11486
	for <policy-archive@odin.ietf.org>; Tue, 27 Jul 1999 11:25:16 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id LAA06628;
	Tue, 27 Jul 1999 11:25:08 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id LAA34982;
	Tue, 27 Jul 1999 11:24:35 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA28356; Tue, 27 Jul 1999 11:08:14 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA09366; Tue, 27 Jul 1999 11:08:09 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id LAA37892
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 11:08:12 -0400
Received: from ckmso1.proxy.att.com (ckmso1.att.com [12.20.58.69])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id LAA37810
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 11:08:04 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by ckmso1.proxy.att.com (AT&T IPNS/MS-2.2) with SMTP id LAA06909;
	Tue, 27 Jul 1999 11:07:39 -0400 (EDT)
Received: from schooner.local.windrose.omaha.ne.us by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id LAA18861; Tue, 27 Jul 1999 11:07:37 -0400
From: "Ryan Moats" <jayhawk@att.com>
To: <ysnir@cisco.com>, "'Yoram Ramberg'" <yramberg@cisco.com>,
        <policy@raleigh.ibm.com>
Subject: RE: ABNF and Versioning (or why Versioning isn't needed)
Date: Tue, 27 Jul 1999 10:05:35 -0500
Message-Id: <002201bed841$7ac8b3c0$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
In-Reply-To: <002901bed842$d0e93c60$475dfe90@ysnir-hpxu.cisco.com>
Importance: Normal
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Ryan Moats" <jayhawk@att.com>
Content-Transfer-Encoding: 7bit

| Please see comments inline.
|
| Yoram Snir
| Cisco Systems
| mailto:\\ysnir@cisco.com
| Tel. 972-9-9700085

[More editing to cut down B/W]

| > [Some editing to cut down B/W]
| >
| > | From: Yoram Ramberg [mailto:yramberg@cisco.com]
| > |
| > | On the switch from SDN to RDN containment:
| > |
| > | This may be tricky. See if I got this one right.
| > | Once you impose strict containment on the DIT you're also presenting a
| > | restriction on relationships between rules and conditions
relationships.
| > | What happens is that your forcing me to place all conditions related
to a
| > | given rule under this rule in the DIT. What if another rule "wants" to
use
| > | one of those conditions? It must copy that condition (if it knows how
to
| > | find it) and place the copy unde the new rule. This makes that
condition
| > | non-reusable. Two identical conditions must be maintained now and
could be
| > | modified independently.
| >
| > Yes and No (and now let me explain).  What the ABNF specifies
| > (and what is the correct way, I might add) is that
| > policyRuleConditionItem and policyRuleActionItem (two new
| > object classes) are restricted to be contained under the
| > policyRule object class.  For policyRuleConditionItem,
| > there are four attributes defined: policyRuleItemGroupNumber,
| > policyRuleItemNegated, policyRuleItemDN, and
| > policyRuleConditionItemName.  The last is a friendly name that
| > MAY be used as an RDN.  The other three are separate attributes
| > that replace the structures of the SDN as separate attributes.
| > Now, while a tuple (group number, negation, DN) [(group number, DN)
| > for policyRuleActionItem] is bound under a single policyRule in
| > the DIT, other policyRules may reuse the DN pointer to build
| > there own policy items.  While it is possible to restructure the
| > ABNF to allow for entire policyRuleConditionItem and
| > policyRuleActionItem
| > reuse, I'm not yet convinced that this is a good thing.  Reusing
| > the DN's achieves 80-90% (my estimate) of full object class reuse
| > without cluttering up the schema with extra DN pointers.
|
| Reusable objects are a very strong mechanism, not only for easier
| definition of policies but also for data integrity, encapsulation,
consistency and
| maintenance of complex policies. Reusable objects are not only conditions
| but variables  and constant values such as the value of HTTP port
| number in a specific domain.
|
| Reusability is used by DiffServ Policers to share bandwidth for example,
| meaning the packets that meet Rule1 and Rule 2 conditions share are marked
| with a certain DSCP value and policed to 64KB/s (together).
| Other examples are in available at
|
ftp://ftpeng.cisco.com/ftp/public/policy/draft-snir-policy-qos-infomodel-00.
txt

Yes, that's the draft I was thinking of in my previous comment "I'm not yet
convinced..."
and yes, that draft talks about doing full reuse of object class instances
through DN
pointers. However, DN pointers are not the only way to go, DIT aliasing is
another
concept that allows for object reuse while maintain the DIT structure [which
I'm
still thinking about].  The point is that using DNs "unless absolutely
necessary"
is adding gratuitous complexity for little or no gain.

| > | (I used conditions in my example because we deem conditions to be
prime
| > | candidate for reusability. Everyone wishing to create a reusable
object
| > | class and place it in the repository will have this problem exactly)
| > |
| > | Furthermore: SDN provide syntax for specifying the intra-expression
| > | relations, i.e., c1 AND c2 and NOT c3 or (NOT c4 abd C7). How would
you
| > | express this if you drop SDNs altogether? We, then, must add
| > | syntax for this kind of expressions. RDN doesn't do the trick
| > | but atructured RDN still may still doesn't solve reusability).
| >
| > The attribute policyRuleItemNegated handles the negation on
| > an object by object basis.
|
| But you need the ability to define a complex arbitrary Boolean expression
| (including parenthesis).

I think there is some confusion.  The original SDN for conditions had
"group:negation:DN".
All that has been done is to replace that attribute with a set of objects
that have
attributes that store "group", "negation" and "DN".  If the original SDN was
rich
enough, the object classes as currently defined are rich enough.

| > SDNs (or sRDNs) are a bad idea because they require the
| > client to process the returned string BEFORE they can
| > determine if they need to make another LDAP query to retrieve
| > the pointed to object.  This means that a client wanting to
| > make only a single LDAP query would be required to pull the ENTIRE
| > policy tree [this is a BAD thing]. Also, by separating out the DN,
| > you are setting up to take advantage of work that is going on
| > in LDAPEXT.  If you use SDNs, you are limited to "rolling your own".
|
| SDN is something that was defined in the CORE policy information model as
a
| tool to define arbitrary, complex boolean expressions. Every alternative
| that may substitude the SDN mechanism will have other complexities since
it
| must support the basic Boolean operators, so a simple lDAP query for
complex
| conditions is problematic. Simple rules can be created by attachment and
| simple LDAP queries could retrieve them in a single query.

While I normally bow to the collect wisdom of a working group, this is an
exception.  I do not see how SDNs "as they are defined" support arbitrary
complex boolean expressions in a way that using the proposed object classes
don't.  Since I have yet to see how the transformation is lossy, if you say
"object classes don't work", I say "then SDNs don't work either".

Ryan



From majordomo@raleigh.ibm.com  Tue Jul 27 12:06:24 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA12802
	for <policy-archive@odin.ietf.org>; Tue, 27 Jul 1999 12:06:24 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA39404;
	Tue, 27 Jul 1999 12:06:16 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id MAA34148;
	Tue, 27 Jul 1999 12:06:13 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA31748; Tue, 27 Jul 1999 11:53:20 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA33012; Tue, 27 Jul 1999 11:53:15 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id LAA34236
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 11:53:17 -0400
Received: from ckmso1.proxy.att.com (ckmso1.att.com [12.20.58.69])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id LAA41106
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 11:53:14 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by ckmso1.proxy.att.com (AT&T IPNS/MS-2.2) with SMTP id LAA29485;
	Tue, 27 Jul 1999 11:52:49 -0400 (EDT)
Received: from schooner.local.windrose.omaha.ne.us by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id LAA02013; Tue, 27 Jul 1999 11:52:47 -0400
From: "Ryan Moats" <jayhawk@att.com>
To: "Grimstad, Al" <AGrimstad@lucentctc.com>
Cc: "Policy Mailing List" <policy@raleigh.ibm.com>
Subject: RE: Long:  Ryan Moat's Promised ABNF for Core Schema Doc
Date: Tue, 27 Jul 1999 10:50:47 -0500
Message-Id: <002801bed847$caf6c480$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
In-Reply-To: <03D3C3E6D4F0D111BBE600104B9470A5012416C3@pigpen.lucentctc.com>
Importance: Normal
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Ryan Moats" <jayhawk@att.com>
Content-Transfer-Encoding: 7bit

| Ryan, it's not that I'm keen on using the RFC2252
| syntax for specifying DEN schemas, but FYI, I did
| notice a small error in your message. When specifying
| DIT content rules, the initial numericoid element is
| that of the structural object class the rule applies to,
| not an independent identifier for the content rule.
| The error occurs for PolicyGroup (3) and PolicyInstance
| (10). -- al

Gurgle...  

Al's right, I missed this in my code (and reviews and ...)

anyway, the correct oids are below:

| >      ( <oid-oc2> NAME 'policyGroup'
| >        DESC 'A container for either a set of related policyRules or a
| >              set of related policyGroups.'
| >        SUP top
| >        MUST (policyGroupName)
| >      )
| > 
| >      ( <oid-cr1>
| >        NAME 'policyGroupContentRule'
| >        DESC 'shows what auxiliary classes go with policyGroup'
| >        AUX (policyGroupContainmentAuxClass $
| > policyRuleContainmentAuxClass)
| >      )
| > 

| In other words, the <oid-cr1> should have been
| <oid-oc2> here.

Al is right <oid-cr1> should be replaced with <oid-oc2>
further, <oid-cr2> should be replaced with <oid-oc11>

Sigh...

Ryan


From majordomo@raleigh.ibm.com  Tue Jul 27 14:44:34 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA17790
	for <policy-archive@odin.ietf.org>; Tue, 27 Jul 1999 14:44:33 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id OAA33348;
	Tue, 27 Jul 1999 14:44:22 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id OAA38590;
	Tue, 27 Jul 1999 14:44:14 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA57658; Tue, 27 Jul 1999 14:33:34 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA52720; Tue, 27 Jul 1999 14:33:27 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id OAA04030
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 14:33:30 -0400
Received: from sol.extremenetworks.com (sol.extremenetworks.com [216.52.8.2])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id OAA31700
	for <policy@raleigh.ibm.com>; Tue, 27 Jul 1999 14:33:26 -0400
Received: by ftp.extremenetworks.com with Internet Mail Service (5.5.2448.0)
	id <36G7J2CY>; Tue, 27 Jul 1999 11:33:06 -0700
Message-Id: <D0805D3B448BD211A7990008C7B181301430E6@ftp.extremenetworks.com>
From: Andrew Smith <andrew@extremenetworks.com>
To: "Policy WG (E-mail)" <policy@raleigh.ibm.com>
Subject: slides and drafts from Oslo?
Date: Tue, 27 Jul 1999 11:33:01 -0700
Mime-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Andrew Smith <andrew@extremenetworks.com>

Could someone post a list of the drafts and pointers to the slides presented
at Oslo Policy WG meeting? Thanks,

Andrew


****************************************************************
Andrew Smith                              tel: +1 (408) 579-2821
Extreme Networks                          fax: +1 (408) 579-3000
3585 Monroe St.                   http://www.extremenetworks.com
Santa Clara CA 95051-1450        em:  andrew@extremenetworks.com
****************************************************************



From majordomo@raleigh.ibm.com  Wed Jul 28 07:35:56 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id HAA04018
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 07:35:54 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id HAA32282;
	Wed, 28 Jul 1999 07:35:40 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id HAA27714;
	Wed, 28 Jul 1999 07:34:47 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA27502; Wed, 28 Jul 1999 07:25:16 -0400
Received: from rtpmail03.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA36238; Wed, 28 Jul 1999 07:24:55 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id HAA26664
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 07:24:58 -0400
Received: from csi-admin1.cisco.com (csi-admin1.cisco.com [144.254.79.12])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id HAA16856
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 07:24:53 -0400
Received: from yramberghpxu (telaviv2-dhcp11.cisco.com [144.254.93.61]) by csi-admin1.cisco.com (8.8.4-Cisco.1/8.6.5) with SMTP id OAA06952; Wed, 28 Jul 1999 14:26:21 +0300 (IDT)
Message-Id: <005701bed92d$fa0a3040$3d5dfe90@cisco.com>
From: "Yoram Ramberg" <yramberg@cisco.com>
To: "Ryan Moats" <jayhawk@att.com>, <ysnir@cisco.com>,
        <policy@raleigh.ibm.com>
References: <002201bed841$7ac8b3c0$e3c8090a@schooner.local.windrose.omaha.ne.us>
Subject: Re: ABNF and Versioning (or why Versioning isn't needed)
Date: Wed, 28 Jul 1999 14:18:29 -0500
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Yoram Ramberg" <yramberg@cisco.com>
Content-Transfer-Encoding: 7bit


(I hope this is, indeed, a general interest subject...)

Ryan, I see your point(s) and it is a good one. We must be held accountable
to the burden our schema design puts on readers, in terms the LDAP roundtrip
count. It is a major optimization issue and we have to resolve this. I'm
glad you're pointing this out at this stage so the schema design may benefit
from this proposed improvement.

The direction toward minimizing DN dereferencing is understood. You're
pointing to a possible confusion in interpreting the expressive power of the
replacement of SDN with three attributes. I'm not confused. In terms of
being able to define an arbitrarily complex Boolean expression, I see no
difference between SDNs and the 3-tuple  (group nr., negation, DN) method.
After all, I can reconstruct all SDNs from their respective 3-tuples, no?
This is mighty fine with me.

However, I still don't understand (shame on me!) how we're going to support
reusability of, say, simple conditions. Let's assume, for the purpose of
this discussion, that reusability is a virtue and we'd like to support it
("bow to the collect wisdom of a working group..."). Please explain how DIT
aliasing allows us to support it. Can we still support the notion of global
repositories? Doesn't this defeat the cause of collecting all related object
in a single subtree so you can pick'em up in one LDAP hit? Don't we end up
paying the same price as in the case of SDNs (if our reusable objects do
reside in a global repository)?

Now, let's see if we can calculate the maximum number of LDAP roundtrips
necessary to fully fetch a complex condition, using the different methods.

SDN (call it Core method):

I think that the number of LDAP queries you'll have to execute to get the
whole
condition depends on the number of instances of simple conditions it is made
of.
It's probably something like:
nr. of simple conditions + 1 (for the Rule)
What does this mean?
A complex condition we'd like to represent is the infamous five tuple IP
header
condition. At a maximum, again, it costs five simple conditions "fetch"
operations to fully gather the darn thing. This added to the hit for the
Rule,
it takes six roundtrips to satisfy us. Remember I said "maximum" twice. I'm
not
convinced this can't be optimized somewhat...

Raju's bulky conditions:

Other methods for modeling conditions in LDAP do get a much better rating on
this. If we follow Raju's schema we find that  same condition (quite
restricted,
though) can be fetched with a single sniff. What the schema pays for that is
measured in its expressive power. Conditions are "bulky" and very
restricted.
To accommodate a slight deviation from the prescribed model you'd have to 1.
Derive your own mechanism (existing classes are too specific to derive
from), 2.
Accept the fact that your rules would be opaque to readers, other than
specially
adapted clients -- your own.

Ryan's 3-tuple method:

In this case the whole complex condition (and, in fact, all complex
conditions!) may be fetched with a single LDAP roundtrip. What's really
attractive here is that we pay no price in terms of expressive power:
Arbitrarily complex conditions can still be specified. What does this method
costs us? I'm yet to fully understand this but I have a hunch that
reusability is the victim, or at least, some reusability. I already asked
Ryan to elaborate on this.

A few words about reusability:

Note that we aren't really talking about object class reusability -- this is
*instance reusability*: The ability to use the same simple condition in
several rules. One reason we want this is pointed out in Yoram Snir's latest
entry in this thread. It's very clear that reusing named condition helps --
we want it (but not at all costs!). Let see if we can clarify this issue by
using another example: Policers.

In the QoS schema there's a concept called "Policer". A Policer models a
certain aggregate treatment to network traffic. Based on our understanding
of QoS Policy requirements and of the DiffServ model we specify three types
of aggregations:

1. Filter aggregation: Treatment given to all traffic that "passes" a given
filter. A rule that has filter made out of 2 simple conditions such as
'(SourcePort == 80) && (sourceIP == 1.2.3.4)' may specify a Policer to set
limits on all traffic from a given WEB server, say, 3Mbps.

2. Class aggregation: DiffServe class, that is. A service class corresponds
to a given DSCP value. If one defines a service class policer, all flows
"colored" by the DSCP of this service class may be aggregated into a single
class-level policer treatment. (We think this is defined in the PHB policy
block)

3. Arbitrary aggregation: We may want to collect several flows into a single
aggregate and police it together. In our current definition we'd just flag a
policer as "shared" and place it in the repository (only repository resident
policers may be "shared"). We now place a reference to this policer in all
rules that contain policy for members of the aggregate. Suppose, for
example, you have two rules that together specify policy for non-urgent
traffic. We want all traffic not related to our help desk application and
our support extranet to be policed together to not ever surpass the 2mbps BW
(under congestion conditions, of course). There are other policies related
to this traffic so it takes several rules to put this thing together:

R1: If (TelnetTraffic) then color = DS1 and Policer is P1;
R2: If (DestinationIP == BigBoss) then color = DS2 and Policer is P1;
R3: If (NOT (HelpDeskApp or SuppExtranetGateway)) then color = DONT_CHANGE
and Policer is P1;

I know this is not perfect but y'all get the point: We may still want to
color flows belonging to the aggregate differently but still want to police
them together.

Using a reusable Policers repository (where policers may be marked 'shared'
(for #3 type aggregation) or 'non-shared' (for #1 type aggregation) allows
us to specify the three behaviors. (The class level policer needs not be
reusable)

I'm assuming that the other benefits of reusable objects are known and
understood -- see Yoram Snir's email for more)

Ok, this is it for now...
*Yoram

----- Original Message -----
From: Ryan Moats <jayhawk@att.com>
To: <ysnir@cisco.com>; 'Yoram Ramberg' <yramberg@cisco.com>;
<policy@raleigh.ibm.com>
Sent: Tuesday, July 27, 1999 10:05 AM
Subject: RE: ABNF and Versioning (or why Versioning isn't needed)


> | Please see comments inline.
> |
> | Yoram Snir
> | Cisco Systems
> | mailto:\\ysnir@cisco.com
> | Tel. 972-9-9700085
>
> [More editing to cut down B/W]
>
> | > [Some editing to cut down B/W]
> | >
> | > | From: Yoram Ramberg [mailto:yramberg@cisco.com]
> | > |
> | > | On the switch from SDN to RDN containment:
> | > |
> | > | This may be tricky. See if I got this one right.
> | > | Once you impose strict containment on the DIT you're also presenting
a
> | > | restriction on relationships between rules and conditions
> relationships.
> | > | What happens is that your forcing me to place all conditions related
> to a
> | > | given rule under this rule in the DIT. What if another rule "wants"
to
> use
> | > | one of those conditions? It must copy that condition (if it knows
how
> to
> | > | find it) and place the copy unde the new rule. This makes that
> condition
> | > | non-reusable. Two identical conditions must be maintained now and
> could be
> | > | modified independently.
> | >
> | > Yes and No (and now let me explain).  What the ABNF specifies
> | > (and what is the correct way, I might add) is that
> | > policyRuleConditionItem and policyRuleActionItem (two new
> | > object classes) are restricted to be contained under the
> | > policyRule object class.  For policyRuleConditionItem,
> | > there are four attributes defined: policyRuleItemGroupNumber,
> | > policyRuleItemNegated, policyRuleItemDN, and
> | > policyRuleConditionItemName.  The last is a friendly name that
> | > MAY be used as an RDN.  The other three are separate attributes
> | > that replace the structures of the SDN as separate attributes.
> | > Now, while a tuple (group number, negation, DN) [(group number, DN)
> | > for policyRuleActionItem] is bound under a single policyRule in
> | > the DIT, other policyRules may reuse the DN pointer to build
> | > there own policy items.  While it is possible to restructure the
> | > ABNF to allow for entire policyRuleConditionItem and
> | > policyRuleActionItem
> | > reuse, I'm not yet convinced that this is a good thing.  Reusing
> | > the DN's achieves 80-90% (my estimate) of full object class reuse
> | > without cluttering up the schema with extra DN pointers.
> |
> | Reusable objects are a very strong mechanism, not only for easier
> | definition of policies but also for data integrity, encapsulation,
> consistency and
> | maintenance of complex policies. Reusable objects are not only
conditions
> | but variables  and constant values such as the value of HTTP port
> | number in a specific domain.
> |
> | Reusability is used by DiffServ Policers to share bandwidth for example,
> | meaning the packets that meet Rule1 and Rule 2 conditions share are
marked
> | with a certain DSCP value and policed to 64KB/s (together).
> | Other examples are in available at
> |
>
ftp://ftpeng.cisco.com/ftp/public/policy/draft-snir-policy-qos-infomodel-00.
> txt
>
> Yes, that's the draft I was thinking of in my previous comment "I'm not
yet
> convinced..."
> and yes, that draft talks about doing full reuse of object class instances
> through DN
> pointers. However, DN pointers are not the only way to go, DIT aliasing is
> another
> concept that allows for object reuse while maintain the DIT structure
[which
> I'm
> still thinking about].  The point is that using DNs "unless absolutely
> necessary"
> is adding gratuitous complexity for little or no gain.
>
> | > | (I used conditions in my example because we deem conditions to be
> prime
> | > | candidate for reusability. Everyone wishing to create a reusable
> object
> | > | class and place it in the repository will have this problem exactly)
> | > |
> | > | Furthermore: SDN provide syntax for specifying the intra-expression
> | > | relations, i.e., c1 AND c2 and NOT c3 or (NOT c4 abd C7). How would
> you
> | > | express this if you drop SDNs altogether? We, then, must add
> | > | syntax for this kind of expressions. RDN doesn't do the trick
> | > | but atructured RDN still may still doesn't solve reusability).
> | >
> | > The attribute policyRuleItemNegated handles the negation on
> | > an object by object basis.
> |
> | But you need the ability to define a complex arbitrary Boolean
expression
> | (including parenthesis).
>
> I think there is some confusion.  The original SDN for conditions had
> "group:negation:DN".
> All that has been done is to replace that attribute with a set of objects
> that have
> attributes that store "group", "negation" and "DN".  If the original SDN
was
> rich
> enough, the object classes as currently defined are rich enough.
>
> | > SDNs (or sRDNs) are a bad idea because they require the
> | > client to process the returned string BEFORE they can
> | > determine if they need to make another LDAP query to retrieve
> | > the pointed to object.  This means that a client wanting to
> | > make only a single LDAP query would be required to pull the ENTIRE
> | > policy tree [this is a BAD thing]. Also, by separating out the DN,
> | > you are setting up to take advantage of work that is going on
> | > in LDAPEXT.  If you use SDNs, you are limited to "rolling your own".
> |
> | SDN is something that was defined in the CORE policy information model
as
> a
> | tool to define arbitrary, complex boolean expressions. Every alternative
> | that may substitude the SDN mechanism will have other complexities since
> it
> | must support the basic Boolean operators, so a simple lDAP query for
> complex
> | conditions is problematic. Simple rules can be created by attachment and
> | simple LDAP queries could retrieve them in a single query.
>
> While I normally bow to the collect wisdom of a working group, this is an
> exception.  I do not see how SDNs "as they are defined" support arbitrary
> complex boolean expressions in a way that using the proposed object
classes
> don't.  Since I have yet to see how the transformation is lossy, if you
say
> "object classes don't work", I say "then SDNs don't work either".
>
> Ryan
>
>



From majordomo@raleigh.ibm.com  Wed Jul 28 08:55:23 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA09859
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 08:55:23 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id IAA36420;
	Wed, 28 Jul 1999 08:55:14 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id IAA32720;
	Wed, 28 Jul 1999 08:55:06 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA44462; Wed, 28 Jul 1999 08:35:10 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA39334; Wed, 28 Jul 1999 08:35:07 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id IAA28302
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 08:35:10 -0400
Received: from wolfe.bbn.com (WOLFE.BBN.COM [128.89.2.232])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id IAA30948
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 08:35:07 -0400
Received: (from mcondell@localhost)
	by wolfe.bbn.com (8.8.8/8.8.6) id IAA05692;
	Wed, 28 Jul 1999 08:35:17 -0400 (EDT)
Date: Wed, 28 Jul 1999 08:35:17 -0400 (EDT)
Message-Id: <199907281235.IAA05692@wolfe.bbn.com>
From: Matthew Condell <mcondell@bbn.com>
To: policy@raleigh.ibm.com
Cc: ipsec-policy@mail.timestep.com
Subject: Policy Framework and IPsec policy
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Matthew Condell <mcondell@bbn.com>


[Sorry for the repeated message to ipsec-policy.  The policy address
 is now spelled correctly]

Hi,

I've been reviewing the policy WG drafts from the perspective
of IPsec policy.  My interest is more in framework issues than
data model/schema issues.  Most of my comments are based on
draft-ietf-policy-terms-02.txt.  

In general, I believe the general policy framework works well 
with the security policy framework, except it one basic assumption.
The general policy framework is based on an intra-domain model
where as IPsec policy MUST deal with inter-domain issues.

However, I think a few modifications to the model would allow
IPsec policy to fit in the framework, without disturbing the
intra-domain model that is already present.

In draft-ietf-policy-terms-02.txt:

Section 4.1:

For inter-domain policy negotiation, different domains will not
share a policy repository, so it is necessary for the PDPs to 
be able to communicate with each other.  We should make sure 
that the architecture does not prohibit this.

Section 5.10:

Inter-domain policy conflicts are not necessarily resolvable, 
since it is not really possible to choose one conflicting 
policy over the other.

Additionally, to allow inter-domain policy caching, it is necessary to
be able to maintain (either in the repository or PDP) policies in a
non-overlapping form. (draft-ietf-ipsec-spp-00.txt refers to
non-overlapping policies as decorrelated.) In decorrelated form, the
conditions of no two policies can be simultaneously satisfied.  The
framework must be able to support such a set of policies to enable it
to support IPsec policy.

This example shows why decorrelation is necessary for an inter-domain
policy framework.  (The example is from the Security Policy Sytem
Architecture draft which is not yet published.)

>Caching correlated policies can lead to incorrect policy
>implementation based on those cached policies, as the following
>example shows.
>
>           H1 --- SG1 --------- SG2 --- H2
>                   |             |  \__ H3
>		    |             |
>		   PS1           PS2
>
>PS2 contains the following policies in its master file:
>	src	dst	proto	direction	action
>   1)	*	H2	*	inbound		permit
>   2)	*	*	*	inbound		deny
>
>These two policies are correlated since all the selectors (src, dst,
>proto, and direction) overlap.  The following SPP exchanges occur:
>
>  1)  PS1 requests policy for H3.
>  2)  PS2 returns policy #2 to PS1 which then caches policy #2.
>  3)  PS1 now looks up the policy for H2 and discovers that it already
>      has a matching policy (policy #2) and uses that.
>
>This is clearly wrong, since policy #2 indicates that the
>communication to H2 should be denied, though PS2's policy actually
>indicates that it should be allowed.
>
>	The solution is to remove the ambiguities that may exist in
>the master file.  The policies of the master file MUST be decorrelated
>before they are entered into the Local Policy Database.  That is, the
>policies must be rewritten so that for every pair of policies there
>exists a selector for which there is a nil intersection between the
>values in both of the policies.
>
>The policies in the above example could be decorrelated as follows:
>	src	dst	proto	direction	action
>   1')	*	H2	*	inbound		permit
>   2')	*	not H2	*	inbound		deny
>
>Now the exchange is a bit different:
>   1)  PS1 requests policy for H3.
>   2)  PS2 returns policy #2' to PS1 which then caches policy #2'.
>   3)  PS1 now looks up the policy for H2, doesn't have a matching 
>       policy, so it requests a policy for H2.
>   4)  PS2 returns policy #1' to PS1 which then caches policy #1', 
>       which is the correct policy for H2.
>

These are the main points where I believe the IPsec policy framework
diverges from general policy framework.

Matthew Condell
BBN Technologies



From majordomo@raleigh.ibm.com  Wed Jul 28 10:58:15 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id KAA15796
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 10:58:14 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id KAA24084;
	Wed, 28 Jul 1999 10:57:50 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id KAA27942;
	Wed, 28 Jul 1999 10:56:54 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA53080; Wed, 28 Jul 1999 10:41:21 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA48594; Wed, 28 Jul 1999 10:41:08 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id KAA38332
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 10:41:11 -0400
Received: from ckmso1.proxy.att.com (ckmso1.att.com [12.20.58.69])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id KAA32368
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 10:41:07 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by ckmso1.proxy.att.com (AT&T IPNS/MS-2.2) with SMTP id KAA17330;
	Wed, 28 Jul 1999 10:40:38 -0400 (EDT)
Received: from schooner.local.windrose.omaha.ne.us by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id KAA10290; Wed, 28 Jul 1999 10:40:35 -0400
From: "Ryan Moats" <jayhawk@att.com>
To: "Yoram Ramberg" <yramberg@cisco.com>, <ysnir@cisco.com>,
        <policy@raleigh.ibm.com>
Subject: RE: ABNF and Versioning (or why Versioning isn't needed)
Date: Wed, 28 Jul 1999 09:38:34 -0500
Message-Id: <001e01bed906$deb222c0$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
In-Reply-To: <005701bed92d$fa0a3040$3d5dfe90@cisco.com>
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Importance: Normal
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Ryan Moats" <jayhawk@att.com>
Content-Transfer-Encoding: 7bit

| (I hope this is, indeed, a general interest subject...)

If it isn't somebody should say so ;-)  On the other hand, the
LDAP schema is a charter item, so it should be ;-).

I'm editing [again] to save BW.  If I leave out an important
point of Yoram's discussion, my apologies in advance.  Anything
I don't reference here, I either agree with or am in small
enough disagreement to not hold it up.

| The direction toward minimizing DN dereferencing is understood. You're
| pointing to a possible confusion in interpreting the expressive
| power of the replacement of SDN with three attributes. I'm not
| confused. In terms of being able to define an arbitrarily complex
| Boolean expression, I see no difference between SDNs and the 3-tuple
| (group nr., negation, DN) method. After all, I can reconstruct all
| SDNs from their respective 3-tuples, no? This is mighty fine with me.

Excellent!  Agreement is a good thing.

|
| However, I still don't understand (shame on me!) how we're going
| to support reusability of, say, simple conditions. Let's assume,
| for the purpose of this discussion, that reusability is a virtue
| and we'd like to support it "bow to the collect wisdom of a working
| group..."). Please explain how DIT aliasing allows us to support it.
| Can we still support the notion of global repositories? Doesn't
| this defeat the cause of collecting all related object in a single
| subtree so you can pick'em up in one LDAP hit? Don't we end up paying
| the same price as in the case of SDNs (if our reusable objects do
| reside in a global repository)?

DIT aliasing (in this discussion) can be viewed as an implementation
detail, so I'd like to take it off the table for the time being.

Well, shame on me for not explaining this clearly (or at all).
Here's my view: the classes policyContion and policyAction are still part
of the schema (as auxiliary classes [hmm, maybe these should be renamed
to conform to typical LDAP schema designs 8-0]) so that we still have
the ability to attach simple conditions or actions to the policyRule class
itself.  Because of this, I don't see where we've lost anything.  In
the case of a simple policy rule, we still attach the aux classes to
the policyRule class [and retrieve in a single LDAP operation].  Note
that this is not reusable [and wasn't in the original schema] in the
sense that an auxiliary class can be attached, but doesn't bring any
default values with it.  The data has to be instantiated separately for
each object in the directory.

This brings up another idea.  If we want, we can consider attaching
the policyCondition and policyAction items to policyRuleConditionItem
and policyRuleActionItem classes (this is currently not allowed in the
ABNF).  This would mean that DNs are only used in the case of very
complex rules that are to be reused.

| Now, let's see if we can calculate the maximum number of LDAP roundtrips
| necessary to fully fetch a complex condition, using the different methods.
|
| SDN (call it Core method):
|
| I think that the number of LDAP queries you'll have to execute to get the
| whole condition depends on the number of instances of simple conditions
| it is made of.
|
| It's probably something like: nr. of simple conditions + 1 (for the Rule)
| What does this mean? A complex condition we'd like to represent is the
| infamous five tuple IP header condition. At a maximum, again, it costs
| five simple conditions "fetch" operations to fully gather the darn thing.
| This added to the hit for the Rule, it takes six roundtrips to satisfy us.
| Remember I said "maximum" twice. I'm not convinced this can't be optimized
| somewhat...
|
| Raju's bulky conditions:
|
| Other methods for modeling conditions in LDAP do get a much better rating
on
| this. If we follow Raju's schema we find that  same condition (quite
| restricted, though) can be fetched with a single sniff. What the schema
pays
| for that is measured in its expressive power. Conditions are "bulky" and
very
| restricted. To accommodate a slight deviation from the prescribed model
you'd
| have to
|
| 1. Derive your own mechanism (existing classes are too specific to derive
| from),
| 2. Accept the fact that your rules would be opaque to readers, other than
| specially adapted clients -- your own.
|
| Ryan's 3-tuple method:
|
| In this case the whole complex condition (and, in fact, all complex
| conditions!) may be fetched with a single LDAP roundtrip. What's really
| attractive here is that we pay no price in terms of expressive power:
| Arbitrarily complex conditions can still be specified. What does
| this method costs us? I'm yet to fully understand this but I have a hunch
that
| reusability is the victim, or at least, some reusability. I already asked
| Ryan to elaborate on this.

This isn't quite right.  While you get all of the conditions in a single
fetch, you don't get the rules pointed to by the DN unless you either

(a) make fetches for each of those rules.  This requires the same number
of fetches as using SDNs.
(b) plan to make use of something like the extended matching rule discussed
in
draft-moats-ldap-dereference-match-00.txt.  This can return everything in
one fetch, but requires support for the extension.
(c) allow for simple policy conditions to be attached to the
policyRuleActionItem
and policyRuleConditionItem classes as discussed above.  This gets
everything
in a single fetch, but sacrifices reusability in the sense that the data
stored
in the auxiliary is not reusable.

My opinion is that the only reusability is the instantiation of the
tuple is not reusable.  The conditions pointed to with the DN are.  By going
with DIT containment and aux classes, you give the policy administrator
freedom:
They can optimize storage to minimize LDAP queries (aux classes) or optimize
for data reuse (DN pointers) or anywhere in between on a policy by policy
basis.
I think this is a good thing.

|
| A few words about reusability:
|
| Note that we aren't really talking about object class reusability
| -- this is *instance reusability*: The ability to use the same simple
| condition in several rules. One reason we want this is pointed out in
Yoram
| Snir's latest entry in this thread. It's very clear that reusing named
| condition helps -- we want it (but not at all costs!). Let see if we can
| clarify this issue by using another example: Policers.

Since, I won't argue the positives of "reusability" and I agree that having
it
"(but not at all costs)" is a good thing, I dropped the rest of the mail
to save BW.  I think that the method (c) above [which will require a delta
to the ABNF schema] is the way to get the best of both worlds in that you
let the administrator make there own decision. I think that is "the best
of both worlds" and I (thinking as a policy administrator) think this is
what I would want.

Seriously though, from my reading, I think we are getting pretty darn
close to some consensus on this (yeah!)

Ryan



From majordomo@raleigh.ibm.com  Wed Jul 28 12:13:12 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA17546
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 12:13:12 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA20408;
	Wed, 28 Jul 1999 12:13:00 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id MAA28470;
	Wed, 28 Jul 1999 12:12:57 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA28488; Wed, 28 Jul 1999 12:00:07 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA55608; Wed, 28 Jul 1999 12:00:03 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id MAA28070
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 12:00:07 -0400
Received: from csi-admin1.cisco.com (csi-admin1.cisco.com [144.254.79.12])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA06854
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 12:00:02 -0400
Received: from cisco.com ([144.254.78.11]) by csi-admin1.cisco.com (8.8.4-Cisco.1/8.6.5) with ESMTP id TAA10667; Wed, 28 Jul 1999 19:01:28 +0300 (IDT)
Message-Id: <379F291E.B2CF645B@cisco.com>
Date: Wed, 28 Jul 1999 19:00:30 +0300
From: Yoram Ramberg <yramberg@cisco.com>
Organization: Cisco Systems Inc.
X-Mailer: Mozilla 4.5 [en] (WinNT; I)
X-Accept-Language: en
Mime-Version: 1.0
To: Ryan Moats <jayhawk@att.com>
Cc: ysnir@cisco.com, policy@raleigh.ibm.com
Subject: Re: ABNF and Versioning (or why Versioning isn't needed)
References: <001e01bed906$deb222c0$e3c8090a@schooner.local.windrose.omaha.ne.us>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Yoram Ramberg <yramberg@cisco.com>
Content-Transfer-Encoding: 7bit

Following the BW frugality trend, I'll do some cruel editing of my own...

Ryan Moats wrote:

> Seriously though, from my reading, I think we are getting pretty darn
> close to some consensus on this (yeah!)

Yes, I think so. I have some evidence for that in the sense that we wish to
allow the freedom of optimization to the user. Note that the QoS schema does
allow for optimization in the case of simple policy - attach and attach and be
done with it. It is not reusable and was not intended to. This is the excuse we
have for most-compact representation. It goes like this:

condition and action are attached to rule.
variable and constant are attached to the condition.
value is attached to constant (always)
possibly: policer & policer are attached to action
Total: One object. Even LDAP requires but a single operation to bring this baby
home.
So we're fine on that end of the stick.

It's arguably the wrong optimization. It makes it very efficient to define
conditions that are *simple*, i.e., an expression of the format '<variable>
<operator> <constant | value>'. It doesn't allow for that arbitrary aggregation
treatment I alluded to before. Obviously not all would agree that this is the
common case. See Raju's schema, for one very different approach to optimization.
We may have to bite this bullet and provide an optimization for some very common
"template" conditions, like that 5-tuple bitch. (sorry)

Are we fine also on the other end? I need to chew on your latest email a bit
more to fully understand it. Maybe Yoram Snir (RDN for the other Yoram) would
have some fresh thoughts...

Thanx,
*Yoram



From majordomo@raleigh.ibm.com  Wed Jul 28 12:16:17 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA17575
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 12:16:16 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA06450;
	Wed, 28 Jul 1999 12:16:01 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id MAA28636;
	Wed, 28 Jul 1999 12:15:59 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA54196; Wed, 28 Jul 1999 12:09:41 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA47372; Wed, 28 Jul 1999 12:09:23 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id MAA21262
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 12:09:26 -0400
Received: from ckmso1.proxy.att.com (ckmso1.att.com [12.20.58.69])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA30186
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 12:09:23 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by ckmso1.proxy.att.com (AT&T IPNS/MS-2.2) with SMTP id MAA01423;
	Wed, 28 Jul 1999 12:09:03 -0400 (EDT)
Received: from schooner.local.windrose.omaha.ne.us by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id MAA25588; Wed, 28 Jul 1999 12:09:01 -0400
From: "Ryan Moats" <jayhawk@att.com>
To: "Yoram Ramberg" <yramberg@cisco.com>
Cc: <ysnir@cisco.com>, <policy@raleigh.ibm.com>
Subject: RE: ABNF and Versioning (or why Versioning isn't needed)
Date: Wed, 28 Jul 1999 11:06:59 -0500
Message-Id: <002401bed913$3912ec20$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
In-Reply-To: <379F291E.B2CF645B@cisco.com>
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Importance: Normal
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Ryan Moats" <jayhawk@att.com>
Content-Transfer-Encoding: 7bit

| Following the BW frugality trend, I'll do some cruel editing of my own...
|
| Ryan Moats wrote:
|
| > Seriously though, from my reading, I think we are getting pretty darn
| > close to some consensus on this (yeah!)
|
| Yes, I think so. I have some evidence for that in the sense that
| we wish to allow the freedom of optimization to the user. Note
| that the QoS schema does allow for optimization in the case of
| simple policy - attach and attach and be done with it. It is not
| reusable and was not intended to. This is the excuse we have for
| most-compact representation. It goes like this:
| condition and action are attached to rule. variable and constant
| are attached to the condition. value is attached to constant (always)
| possibly: policer & policer are attached to action Total: One object.
| Even LDAP requires but a single operation to bring this baby home.
| So we're fine on that end of the stick.
|
| It's arguably the wrong optimization. It makes it very efficient
| to define conditions that are *simple*, i.e., an expression of the
| format '<variable> <operator> <constant | value>'. It doesn't allow
| for that arbitrary aggregation treatment I alluded to before. Obviously
| not all would agree that this is the common case. See Raju's schema,
| for one very different approach to optimization. We may have to bite
| this bullet and provide an optimization for some very common
| "template" conditions, like that 5-tuple bitch. (sorry)

Well, I think that my suggestion (in previous email) of allowing
policyCondition
and  policyAction auxiliary classes to be attached to
policyRuleConditionItem
and policyRuleActionItem, respectively solves this.  But definitely
chew on it some more (cause I could be mistaken)

| Are we fine also on the other end? I need to chew on your latest
| email a bit more to fully understand it. Maybe Yoram Snir (RDN for
| the other Yoram) would have some fresh thoughts...

I think we are REAL close (the more I think about it, the more
I think that the aux class attachment method to the child
classes is a GOOD thing and should be added).

Ryan



From majordomo@raleigh.ibm.com  Wed Jul 28 13:26:56 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA19254
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 13:26:56 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA27950;
	Wed, 28 Jul 1999 13:26:45 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id NAA37688;
	Wed, 28 Jul 1999 13:26:41 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA39518; Wed, 28 Jul 1999 13:14:40 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA46398; Wed, 28 Jul 1999 13:14:12 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id NAA38192
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 13:14:16 -0400
Received: from smtprtp1.nortel.net (nrtcom-137-118-22-14.ntcom.nortel.net [137.118.22.14])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA17768
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 13:14:12 -0400
Received: from zrchb213.us.nortel.com (actually zrchb213) 
          by smtprtp1.nortel.net; Wed, 28 Jul 1999 13:13:17 -0400
Received: from zrtpd003.us.nortel.com ([47.140.224.137]) 
          by zrchb213.us.nortel.com 
          with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2448.0) 
          id PYN10R7T; Wed, 28 Jul 1999 12:13:15 -0500
Received: from americasm01.nt.com (wppkhc0b.us.nortel.com [47.39.57.22]) 
          by zrtpd003.us.nortel.com 
          with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2448.0) 
          id P3LNXN4M; Wed, 28 Jul 1999 13:13:17 -0400
Message-Id: <379F3A2B.9AC82B6D@americasm01.nt.com>
Date: Wed, 28 Jul 1999 13:13:15 -0400
From: "He Huang" <huanghe@nortelnetworks.com>
Organization: Nortel Networks
X-Mailer: Mozilla 4.06 [en] (X11; U; HP-UX B.10.20 9000/712)
Mime-Version: 1.0
To: Matthew Condell <mcondell@bbn.com>, policy@raleigh.ibm.com,
        ipsec-policy@mail.timestep.com
Subject: Re: Policy Framework and IPsec policy
References: <199907281235.IAA05692@wolfe.bbn.com>
Content-Type: text/plain; charset=big5
Content-Transfer-Encoding: 7bit
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "He Huang" <huanghe@nortelnetworks.com>
Content-Transfer-Encoding: 7bit

Hi, 

Certainly, before one policy is saved in policy repository/master
database, it's necessary to do conflict detection and resolution. The
question is that who is responsible for that work, PolicyFramework or
LDAP itself(Assume repository is LDAP directory). In addition, Policy
Framework suggests two solutions for policy conflict(
draft-strassner-policy- terms-00.txt): first matching rule and priority
rule. It seems not to work well in some situations. Any new solution for
policy conflict?

Matthew Condell wrote:
> 
> [Sorry for the repeated message to ipsec-policy.  The policy address
>  is now spelled correctly]
> 
> Hi,
> 
> I've been reviewing the policy WG drafts from the perspective
> of IPsec policy.  My interest is more in framework issues than
> data model/schema issues.  Most of my comments are based on
> draft-ietf-policy-terms-02.txt.
> 
> In general, I believe the general policy framework works well
> with the security policy framework, except it one basic assumption.
> The general policy framework is based on an intra-domain model
> where as IPsec policy MUST deal with inter-domain issues.
> 
> However, I think a few modifications to the model would allow
> IPsec policy to fit in the framework, without disturbing the
> intra-domain model that is already present.
> 
> In draft-ietf-policy-terms-02.txt:
> 
> Section 4.1:
> 
> For inter-domain policy negotiation, different domains will not
> share a policy repository, so it is necessary for the PDPs to
> be able to communicate with each other.  We should make sure
> that the architecture does not prohibit this.
> 
> Section 5.10:
> 
> Inter-domain policy conflicts are not necessarily resolvable,
> since it is not really possible to choose one conflicting
> policy over the other.
> 
> Additionally, to allow inter-domain policy caching, it is necessary to
> be able to maintain (either in the repository or PDP) policies in a
> non-overlapping form. (draft-ietf-ipsec-spp-00.txt refers to
> non-overlapping policies as decorrelated.) In decorrelated form, the
> conditions of no two policies can be simultaneously satisfied.  The
> framework must be able to support such a set of policies to enable it
> to support IPsec policy.
> 
> This example shows why decorrelation is necessary for an inter-domain
> policy framework.  (The example is from the Security Policy Sytem
> Architecture draft which is not yet published.)
> 
> >Caching correlated policies can lead to incorrect policy
> >implementation based on those cached policies, as the following
> >example shows.
> >
> >           H1 --- SG1 --------- SG2 --- H2
> >                   |             |  \__ H3
> >                   |             |
> >                  PS1           PS2
> >
> >PS2 contains the following policies in its master file:
> >       src     dst     proto   direction       action
> >   1)  *       H2      *       inbound         permit
> >   2)  *       *       *       inbound         deny
> >
> >These two policies are correlated since all the selectors (src, dst,
> >proto, and direction) overlap.  The following SPP exchanges occur:
> >
> >  1)  PS1 requests policy for H3.
> >  2)  PS2 returns policy #2 to PS1 which then caches policy #2.
> >  3)  PS1 now looks up the policy for H2 and discovers that it already
> >      has a matching policy (policy #2) and uses that.
> >
> >This is clearly wrong, since policy #2 indicates that the
> >communication to H2 should be denied, though PS2's policy actually
> >indicates that it should be allowed.
> >
> >       The solution is to remove the ambiguities that may exist in
> >the master file.  The policies of the master file MUST be decorrelated
> >before they are entered into the Local Policy Database.  That is, the
> >policies must be rewritten so that for every pair of policies there
> >exists a selector for which there is a nil intersection between the
> >values in both of the policies.
> >
> >The policies in the above example could be decorrelated as follows:
> >       src     dst     proto   direction       action
> >   1') *       H2      *       inbound         permit
> >   2') *       not H2  *       inbound         deny
> >
> >Now the exchange is a bit different:
> >   1)  PS1 requests policy for H3.
> >   2)  PS2 returns policy #2' to PS1 which then caches policy #2'.
> >   3)  PS1 now looks up the policy for H2, doesn't have a matching
> >       policy, so it requests a policy for H2.
> >   4)  PS2 returns policy #1' to PS1 which then caches policy #1',
> >       which is the correct policy for H2.
> >
> 
> These are the main points where I believe the IPsec policy framework
> diverges from general policy framework.
> 
> Matthew Condell
> BBN Technologies


From majordomo@raleigh.ibm.com  Wed Jul 28 13:59:14 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id NAA19870
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 13:59:13 -0400 (EDT)
Received: from rtpmail01.raleigh.ibm.com (rtpmail01.raleigh.ibm.com [9.37.172.24])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA43206;
	Wed, 28 Jul 1999 13:59:03 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id NAB23580;
	Wed, 28 Jul 1999 13:59:01 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA54962; Wed, 28 Jul 1999 13:46:58 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA46506; Wed, 28 Jul 1999 13:46:55 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id NAA37788
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 13:46:49 -0400
Received: from wolfe.bbn.com (WOLFE.BBN.COM [128.89.2.232])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id NAA29446
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 13:46:40 -0400
Received: (from mcondell@localhost)
	by wolfe.bbn.com (8.8.8/8.8.6) id NAA07663;
	Wed, 28 Jul 1999 13:46:40 -0400 (EDT)
Date: Wed, 28 Jul 1999 13:46:40 -0400 (EDT)
Message-Id: <199907281746.NAA07663@wolfe.bbn.com>
From: Matthew Condell <mcondell@bbn.com>
To: "He Huang" <huanghe@nortelnetworks.com>
Cc: policy@raleigh.ibm.com, ipsec-policy@mail.timestep.com
References: <379F3A2B.9AC82B6D@americasm01.nt.com>
Subject: Re: Policy Framework and IPsec policy
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Matthew Condell <mcondell@bbn.com>


Hi, 

>Certainly, before one policy is saved in policy repository/master
>database, it's necessary to do conflict detection and resolution. The
>question is that who is responsible for that work, PolicyFramework or
>LDAP itself(Assume repository is LDAP directory). In addition, Policy
>Framework suggests two solutions for policy conflict(
>draft-strassner-policy- terms-00.txt): first matching rule and priority
>rule. It seems not to work well in some situations. Any new solution for
>policy conflict?

There are really two levels of conflict resolution when dealing with
inter-domain policy.  The intra-domain conflict resolution which can
be solved using first match or priorities.  

Inter-domain conflict resolution is different, however.  If two
policies that are from different domains conflict about a particular
communication, there is no appropriate resolution since the two
policies are established by two different administrative domains. The
communication may just not be possible.  The framework should
acknowledge this possiblility to not exclude inter-domain policy from
fitting withing the framework.

Matthew Condell
BBN Technologies


From majordomo@raleigh.ibm.com  Wed Jul 28 14:22:48 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id OAA20409
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 14:22:48 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id OAA32402;
	Wed, 28 Jul 1999 14:22:29 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id OAA28320;
	Wed, 28 Jul 1999 14:22:23 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA08270; Wed, 28 Jul 1999 14:10:52 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA08222; Wed, 28 Jul 1999 14:10:48 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id OAA27024
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 14:10:50 -0400
Received: from alms1.fw.att.com (alms1.att.com [192.128.167.146])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id OAA30600
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 14:10:46 -0400
Received: from qsun.ho.att.com ([135.16.30.2])
	by alms1.fw.att.com (AT&T IPNS/MS-2.2) with SMTP id OAA00742;
	Wed, 28 Jul 1999 14:10:21 -0400 (EDT)
Received: from schooner.local.windrose.omaha.ne.us by qsun.ho.att.com (SMI-8.6/EMS-1.2 sol2)
	id OAA17595; Wed, 28 Jul 1999 14:10:19 -0400
From: "Ryan Moats" <jayhawk@att.com>
To: "Matthew Condell" <mcondell@bbn.com>,
        "He Huang" <huanghe@nortelnetworks.com>
Cc: <policy@raleigh.ibm.com>, <ipsec-policy@mail.timestep.com>
Subject: RE: Policy Framework and IPsec policy
Date: Wed, 28 Jul 1999 13:08:16 -0500
Message-Id: <003001bed924$2a94d6c0$e3c8090a@schooner.local.windrose.omaha.ne.us>
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3 (Normal)
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook 8.5, Build 4.71.2377.0
In-Reply-To: <199907281746.NAA07663@wolfe.bbn.com>
X-Mimeole: Produced By Microsoft MimeOLE V4.72.3155.0
Importance: Normal
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Ryan Moats" <jayhawk@att.com>
Content-Transfer-Encoding: 7bit

| Hi, 
| 
| >Certainly, before one policy is saved in policy repository/master
| >database, it's necessary to do conflict detection and resolution. The
| >question is that who is responsible for that work, PolicyFramework or
| >LDAP itself(Assume repository is LDAP directory). In addition, Policy
| >Framework suggests two solutions for policy conflict(
| >draft-strassner-policy- terms-00.txt): first matching rule and priority
| >rule. It seems not to work well in some situations. Any new solution for
| >policy conflict?
| 
| There are really two levels of conflict resolution when dealing with
| inter-domain policy.  The intra-domain conflict resolution which can
| be solved using first match or priorities.  
| 
| Inter-domain conflict resolution is different, however.  If two
| policies that are from different domains conflict about a particular
| communication, there is no appropriate resolution since the two
| policies are established by two different administrative domains. The
| communication may just not be possible.  The framework should
| acknowledge this possiblility to not exclude inter-domain policy from
| fitting withing the framework.

Just a technology note.  LDAP is not going to do conflict resolution
beyond anything provided by LDUP (replication conflict resolution).
Conflict resolution that involves syntactical interpretation of the data
is not something LDAP does (or was ever intended to do).

Ryan


From majordomo@raleigh.ibm.com  Wed Jul 28 22:47:17 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id WAA28455
	for <policy-archive@odin.ietf.org>; Wed, 28 Jul 1999 22:47:17 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id WAA24250;
	Wed, 28 Jul 1999 22:47:08 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id WAA26416;
	Wed, 28 Jul 1999 22:47:03 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA55030; Wed, 28 Jul 1999 22:37:03 -0400
Received: from rtpmail01.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA34106; Wed, 28 Jul 1999 22:36:46 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail01.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id WAA12056
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 22:36:50 -0400
Received: from bmailnj.iphighway.com ([209.3.6.76])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id WAA26348
	for <policy@raleigh.ibm.com>; Wed, 28 Jul 1999 22:36:49 -0400
Received: from shai-notebook (slip-32-100-243-195.ma.us.IBM.NET [32.100.243.195]) by bmailnj.iphighway.com with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2448.0)
	id PQDW91AC; Wed, 28 Jul 1999 22:28:41 -0400
Message-Id: <4.2.0.58.19990728120350.00a94e90@bmailnj.iphighway.com>
X-Sender: herzog@bmailnj.iphighway.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.2.0.58 
Date: Wed, 28 Jul 1999 12:11:29 -0400
To: stevej@DeterministicNetworks.com (SteveJ),
        "Andrew Smith" <andrew@extremenetworks.com>,
        "'Pradosh Kumar Mohapatra'" <mohap@research.bell-labs.com>
From: Shai Herzog <herzog@iphighway.com>
Subject: Re: COPS message content
Cc: "rap (E-mail)" <rap@iphighway.com>, "policy" <policy@raleigh.ibm.com>
In-Reply-To: <001501bed25c$fa74dee0$da67b3cd@DeterministicNetworks.com>
References: <D0805D3B448BD211A7990008C7B1813014306F@ftp.extremenetworks.com>
Mime-Version: 1.0
Content-Type: multipart/alternative;
	boundary="=====================_24389670==_.ALT"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Shai Herzog <herzog@iphighway.com>

--=====================_24389670==_.ALT
Content-Type: text/plain; charset="us-ascii"; format=flowed

At 02:07 PM 7/19/99, SteveJ wrote:

>I believe
>the jury is still out on how the Policy Server/Repository -- PDP -- PEP will
>actually be implemented.

I always believed Policy Server and PDP are the same thing so the three
tier architecture is

Repository -- PDP -- PEP

And the protocol used are LDAP (Repository -- PDP) and COPS (PDP-PEP).

Shai

__________________________________________________________________
Shai Herzog, Founder & CTO         http://www.iphighway.com/herzog
Cell: (917) 449-7889                      Parker Plaza, 16th Floor
Tel : (201) 585-0800                                 400 Kelby St.
Fax : (212) 656-1006                             Fort-Lee NJ 07024 
--=====================_24389670==_.ALT
Content-Type: text/html; charset="us-ascii"

<html>
At 02:07 PM 7/19/99, SteveJ wrote:<br>
<br>
<blockquote type=cite cite>I believe<br>
the jury is still out on how the Policy Server/Repository -- PDP -- PEP
will<br>
actually be implemented.&nbsp; </blockquote><br>
I always believed Policy Server and PDP are the same thing so the
three<br>
tier architecture is<br>
<br>
Repository -- PDP -- PEP<br>
<br>
And the protocol used are LDAP (Repository -- PDP) and COPS
(PDP-PEP).<br>
<br>
Shai<br>
<br>
<div>__________________________________________________________________</div>
<div>Shai Herzog, Founder &amp;
CTO&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<a href="http://www.iphighway.com/herzog" EUDORA=AUTOURL>http://www.iphighway.com/herzog</a></div>
<div>Cell: (917)
449-7889&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Parker Plaza, 16th Floor</div>
<div>Tel : (201)
585-0800&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
400 Kelby St.</div>
Fax : (212)
656-1006&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Fort-Lee NJ 07024
</html>

--=====================_24389670==_.ALT--



From majordomo@raleigh.ibm.com  Thu Jul 29 11:04:01 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA23204
	for <policy-archive@odin.ietf.org>; Thu, 29 Jul 1999 11:03:54 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id LAA18066;
	Thu, 29 Jul 1999 11:03:06 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id LAA34918;
	Thu, 29 Jul 1999 11:02:40 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA38658; Thu, 29 Jul 1999 10:47:32 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA33018; Thu, 29 Jul 1999 10:47:29 -0400
Received: from fwns2.raleigh.ibm.com (fwns2.raleigh.ibm.com [9.37.0.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id KAA36404
	for <policy@raleigh.ibm.com>; Thu, 29 Jul 1999 10:47:31 -0400
Received: from apollo.sitaranetworks.com (apollo.sitaranetworks.com [199.103.141.105])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id KAA39412
	for <policy@raleigh.ibm.com>; Thu, 29 Jul 1999 10:47:03 -0400
Message-Id: <001601bed9d1$a0758080$728d67c7@sitaranetworks.com>
From: "Vladimir Sukonnik" <vladimir@sitaranetworks.com>
To: <policy@raleigh.ibm.com>
Subject: status of PFDL
Date: Thu, 29 Jul 1999 10:49:57 -0400
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Vladimir Sukonnik" <vladimir@sitaranetworks.com>
Content-Transfer-Encoding: 7bit

What's the status of this document? Has anyone implemented a parser for the
BNF proposed at the end of the draft?

Thank you,
V

Vladimir Sukonnik
Sitara Networks, Inc.



From majordomo@raleigh.ibm.com  Thu Jul 29 16:01:50 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id QAA02222
	for <policy-archive@odin.ietf.org>; Thu, 29 Jul 1999 16:01:49 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id QAA30218;
	Thu, 29 Jul 1999 16:01:39 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id QAA27386;
	Thu, 29 Jul 1999 16:01:26 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA53112; Thu, 29 Jul 1999 15:46:29 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA33154; Thu, 29 Jul 1999 15:46:04 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id PAA36430
	for <policy@raleigh.ibm.com>; Thu, 29 Jul 1999 15:46:06 -0400
Received: from diablo.cisco.com (diablo.cisco.com [171.68.224.210])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id PAA26276
	for <policy@raleigh.ibm.com>; Thu, 29 Jul 1999 15:46:04 -0400
Received: from jschnizl1-pc.cisco.com (jschnizl-isdn.cisco.com [171.70.238.115]) by diablo.cisco.com (8.8.6 (PHNE_14041)/CISCO.SERVER.1.2) with SMTP id MAA27170; Thu, 29 Jul 1999 12:45:36 -0700 (PDT)
Message-Id: <4.1.19990729125236.00aa22d0@diablo.cisco.com>
X-Sender: jschnizl@diablo.cisco.com
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 
Date: Thu, 29 Jul 1999 15:45:06 -0400
To: "He Huang" <huanghe@nortelnetworks.com>,
        Matthew Condell <mcondell@bbn.com>, policy@raleigh.ibm.com,
        ipsec-policy@mail.timestep.com
From: John Schnizlein <jschnizl@cisco.com>
Subject: Re: Policy Framework and IPsec policy
In-Reply-To: <379F3A2B.9AC82B6D@americasm01.nt.com>
References: <199907281235.IAA05692@wolfe.bbn.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: John Schnizlein <jschnizl@cisco.com>

At 01:13 PM 07/28/1999 -0400, He Huang wrote:
>
>Certainly, before one policy is saved in policy repository/master
>database, it's necessary to do conflict detection and resolution. The
>question is that who is responsible for that work, PolicyFramework or
>LDAP itself (Assume repository is LDAP directory)...

This is not the question. An LDAP (repository) cannot be responsible
for detecting and resolving conflicts in the semantics of a policy.
Its capability is limited to detecting conflicts (most would call 
these attribute or syntax violation rather than conflicts) between
input attributes/values and the schema.

Matthew's point is that there are potential semantic conflicts between
different administrative domains even after all internal conflicts are
resolved in each domain. The assumption is that some negotiation is
necessary between the domains. If I understand his proposal, this
negotiation is easier if those rules that might conflict (are relevant
to the other domain) are separated and disambiguated with those of the
other domain as early in the process as possible.

Even though this proposal requires an explicit negotiation cycle, in
addition to a distinction in the policy store we have not considered
in the policy framework WG, is much more attractive than surprises
when the execution of policy is other than what was intended.

Have security folks determined if there is sufficient value in keeping
policy for different domains in separate subtrees of the DIT?

John


From majordomo@raleigh.ibm.com  Fri Jul 30 09:54:50 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA02847
	for <policy-archive@odin.ietf.org>; Fri, 30 Jul 1999 09:54:49 -0400 (EDT)
Received: from rtpmail03.raleigh.ibm.com (rtpmail03.raleigh.ibm.com [9.37.172.47])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id JAA28320;
	Fri, 30 Jul 1999 09:54:34 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail03.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id JAA27376;
	Fri, 30 Jul 1999 09:54:19 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA55402; Fri, 30 Jul 1999 09:33:55 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA55300; Fri, 30 Jul 1999 09:33:44 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id JAA39528
	for <policy@raleigh.ibm.com>; Fri, 30 Jul 1999 09:33:47 -0400
Received: from alcor.process.com (alcor.process.com [192.42.95.16])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id JAA16654
	for <policy@raleigh.ibm.com>; Fri, 30 Jul 1999 09:33:44 -0400
Received: from bvolz (192.42.95.238) by alcor.process.com (MX V5.1-X A2w8g)
          with SMTP; Fri, 30 Jul 1999 09:33:20 -0400
Message-Id: <000901beda8f$b9fd0d20$69e98018@process.com>
From: "Bernie Volz" <volz@process.com>
To: "Matthew Condell" <mcondell@BBN.COM>, <policy@raleigh.ibm.com>
Cc: <ipsec-policy@mail.timestep.com>
References: <199907281235.IAA05692@wolfe.bbn.com>
Subject: Re: Policy Framework and IPsec policy
Date: Fri, 30 Jul 1999 09:30:43 -0400
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "Bernie Volz" <volz@process.com>
Content-Transfer-Encoding: 7bit

FYI - Have implemented NFS (and other caching protocols) on platforms that
were not native to NFS, I have one piece of advice for doing caching. It is
best to use cached data ONLY when the question is the same. So, in your
examples, PS1 could only use the cached data (that it got for H3) if it was
again trying to see what the policies were for H3. It could not use it if it
was asking what the policies where for H1. The more complex a system of
conditions you have, the more important this becomes. I feel the PS<n> can
not assume that information it got about one question applies to a
completely different question. This does make caching less useful (and
requires the cache to store much more), but it really is the only safe way
to go, IMHO.

It also avoids your problem (though it doesn't resolve the issue of
conflicting policies in separate administrative domains).

- Bernie Volz
  Process Software

----- Original Message -----
From: Matthew Condell <mcondell@BBN.COM>
To: <policy@raleigh.ibm.com>
Cc: <ipsec-policy@mail.timestep.com>
Sent: Wednesday, July 28, 1999 8:35 AM
Subject: Policy Framework and IPsec policy


>
> [Sorry for the repeated message to ipsec-policy.  The policy address
>  is now spelled correctly]
>
> Hi,
>
> I've been reviewing the policy WG drafts from the perspective
> of IPsec policy.  My interest is more in framework issues than
> data model/schema issues.  Most of my comments are based on
> draft-ietf-policy-terms-02.txt.
>
> In general, I believe the general policy framework works well
> with the security policy framework, except it one basic assumption.
> The general policy framework is based on an intra-domain model
> where as IPsec policy MUST deal with inter-domain issues.
>
> However, I think a few modifications to the model would allow
> IPsec policy to fit in the framework, without disturbing the
> intra-domain model that is already present.
>
> In draft-ietf-policy-terms-02.txt:
>
> Section 4.1:
>
> For inter-domain policy negotiation, different domains will not
> share a policy repository, so it is necessary for the PDPs to
> be able to communicate with each other.  We should make sure
> that the architecture does not prohibit this.
>
> Section 5.10:
>
> Inter-domain policy conflicts are not necessarily resolvable,
> since it is not really possible to choose one conflicting
> policy over the other.
>
> Additionally, to allow inter-domain policy caching, it is necessary to
> be able to maintain (either in the repository or PDP) policies in a
> non-overlapping form. (draft-ietf-ipsec-spp-00.txt refers to
> non-overlapping policies as decorrelated.) In decorrelated form, the
> conditions of no two policies can be simultaneously satisfied.  The
> framework must be able to support such a set of policies to enable it
> to support IPsec policy.
>
> This example shows why decorrelation is necessary for an inter-domain
> policy framework.  (The example is from the Security Policy Sytem
> Architecture draft which is not yet published.)
>
> >Caching correlated policies can lead to incorrect policy
> >implementation based on those cached policies, as the following
> >example shows.
> >
> >           H1 --- SG1 --------- SG2 --- H2
> >                   |             |  \__ H3
> >     |             |
> >    PS1           PS2
> >
> >PS2 contains the following policies in its master file:
> > src dst proto direction action
> >   1) * H2 * inbound permit
> >   2) * * * inbound deny
> >
> >These two policies are correlated since all the selectors (src, dst,
> >proto, and direction) overlap.  The following SPP exchanges occur:
> >
> >  1)  PS1 requests policy for H3.
> >  2)  PS2 returns policy #2 to PS1 which then caches policy #2.
> >  3)  PS1 now looks up the policy for H2 and discovers that it already
> >      has a matching policy (policy #2) and uses that.
> >
> >This is clearly wrong, since policy #2 indicates that the
> >communication to H2 should be denied, though PS2's policy actually
> >indicates that it should be allowed.
> >
> > The solution is to remove the ambiguities that may exist in
> >the master file.  The policies of the master file MUST be decorrelated
> >before they are entered into the Local Policy Database.  That is, the
> >policies must be rewritten so that for every pair of policies there
> >exists a selector for which there is a nil intersection between the
> >values in both of the policies.
> >
> >The policies in the above example could be decorrelated as follows:
> > src dst proto direction action
> >   1') * H2 * inbound permit
> >   2') * not H2 * inbound deny
> >
> >Now the exchange is a bit different:
> >   1)  PS1 requests policy for H3.
> >   2)  PS2 returns policy #2' to PS1 which then caches policy #2'.
> >   3)  PS1 now looks up the policy for H2, doesn't have a matching
> >       policy, so it requests a policy for H2.
> >   4)  PS2 returns policy #1' to PS1 which then caches policy #1',
> >       which is the correct policy for H2.
> >
>
> These are the main points where I believe the IPsec policy framework
> diverges from general policy framework.
>
> Matthew Condell
> BBN Technologies
>
>



From majordomo@raleigh.ibm.com  Fri Jul 30 12:38:26 1999
Received: from fwns2.raleigh.ibm.com (fwns2d.raleigh.ibm.com [204.146.167.236])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id MAA06540
	for <policy-archive@odin.ietf.org>; Fri, 30 Jul 1999 12:38:24 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns2.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id MAA36748;
	Fri, 30 Jul 1999 12:38:09 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id MAA26400;
	Fri, 30 Jul 1999 12:37:38 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA30618; Fri, 30 Jul 1999 10:52:45 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA45910; Fri, 30 Jul 1999 10:52:37 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id KAA38092
	for <policy@raleigh.ibm.com>; Fri, 30 Jul 1999 10:52:39 -0400
Received: from wolfe.bbn.com (WOLFE.BBN.COM [128.89.2.232])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id KAA27724
	for <policy@raleigh.ibm.com>; Fri, 30 Jul 1999 10:52:37 -0400
Received: (from mcondell@localhost)
	by wolfe.bbn.com (8.8.8/8.8.6) id KAA18980;
	Fri, 30 Jul 1999 10:52:49 -0400 (EDT)
Date: Fri, 30 Jul 1999 10:52:49 -0400 (EDT)
Message-Id: <199907301452.KAA18980@wolfe.bbn.com>
From: Matthew Condell <mcondell@bbn.com>
To: policy@raleigh.ibm.com
Cc: ipsec-policy@mail.timestep.com
References: <000901beda8f$b9fd0d20$69e98018@process.com>
Subject: Re: Policy Framework and IPsec policy
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: Matthew Condell <mcondell@bbn.com>


Caching policies from another server is actually safe, if you can
guarantee that all the servers keep their policies in a decorrelated
form.  (More on this below.)  The system we have for exchanging IPsec
policies requires all policies to be decorrelated, so our policy
servers *can* assume that information it got about one question
applies to a completely different question, if the conditions match.

Decorrelation is a process that removes all overlap of the conditions
in a set of policy rules.  The process maintains the implicit
information about the policies that is contained in the rule ordering
(using either first match or priorities).  As a result, a question
can only match the conditions of *one* policy rule.  Also there is no
longer any need to order the policies.

Decorrelation is explained in more detail in draft-ietf-ipsec-sps-00.txt.  
An algorithm for decorrelation is also presented there.  (The draft
has expired, but is still in the drafts repository.  The draft is
currently being reworked and split into multiple drafts.  The new
draft with the decorrelation process is still in the works.)

My main concern here is to make sure that the policy framework doesn't
prohibit policies to be stored in the repository in a decorrelated
form so that an inter-domain policy protocol that requires
decorrelation will still fit into the gereral policy framework.

Matthew Condell
BBN Technologies

>FYI - Have implemented NFS (and other caching protocols) on platforms that
>were not native to NFS, I have one piece of advice for doing caching. It is
>best to use cached data ONLY when the question is the same. So, in your
>examples, PS1 could only use the cached data (that it got for H3) if it was
>again trying to see what the policies were for H3. It could not use it if it
>was asking what the policies where for H1. The more complex a system of
>conditions you have, the more important this becomes. I feel the PS<n> can
>not assume that information it got about one question applies to a
>completely different question. This does make caching less useful (and
>requires the cache to store much more), but it really is the only safe way
>to go, IMHO.


From majordomo@raleigh.ibm.com  Fri Jul 30 17:58:17 1999
Received: from fwns1.raleigh.ibm.com (fwns1d.raleigh.ibm.com [204.146.167.235])
	by ietf.org (8.9.1a/8.9.1a) with ESMTP id RAA10390
	for <policy-archive@odin.ietf.org>; Fri, 30 Jul 1999 17:58:16 -0400 (EDT)
Received: from rtpmail02.raleigh.ibm.com (rtpmail02.raleigh.ibm.com [9.37.172.48])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id RAA22444;
	Fri, 30 Jul 1999 17:58:05 -0400
Received: from rtpaix11.raleigh.ibm.com (rtpaix11.raleigh.ibm.com [9.37.172.4])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with SMTP id RAA32262;
	Fri, 30 Jul 1999 17:57:56 -0400
Received: by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA56434; Fri, 30 Jul 1999 17:38:03 -0400
Received: from rtpmail02.raleigh.ibm.com by rtpaix11.raleigh.ibm.com (AIX 4.1/UCB 5.64/4.03-RAL)
          id AA30314; Fri, 30 Jul 1999 17:38:00 -0400
Received: from fwns1.raleigh.ibm.com (fwns1.raleigh.ibm.com [9.37.0.3])
	by rtpmail02.raleigh.ibm.com (8.8.5/8.8.5/RTP-ral-1.1) with ESMTP id RAA20572
	for <POLICY@raleigh.ibm.com>; Fri, 30 Jul 1999 17:38:04 -0400
Received: from omega.cisco.com (omega.cisco.com [171.69.63.141])
	by fwns1.raleigh.ibm.com (8.9.0/8.9.0/RTP-FW-1.2) with ESMTP id RAA06788
	for <POLICY@raleigh.ibm.com>; Fri, 30 Jul 1999 17:37:08 -0400
Received: from jstrassnlap.cisco.com ([171.69.108.130])
	by omega.cisco.com (8.8.8-Cisco List Logging/8.8.8) with SMTP id OAA06259;
	Fri, 30 Jul 1999 14:36:46 -0700 (PDT)
Message-Id: <04ea01bedad3$e0a029a0$826c45ab@cisco.com>
From: "John Strassner" <jstrassn@cisco.com>
To: <POLICY@raleigh.ibm.com>
Cc: <ellesson@us.ibm.com>
Subject: Draft Policy minutes
Date: Fri, 30 Jul 1999 14:38:34 -0700
Organization: Cisco Systems
Mime-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-Msmail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-Mimeole: Produced By Microsoft MimeOLE V5.00.2314.1300
Sender: policy-owner@raleigh.ibm.com
Precedence: bulk
Reply-To: "John Strassner" <jstrassn@cisco.com>
Content-Transfer-Encoding: 7bit

Policy Enthusiasts,

attached please find the draft minutes for the Policy Framework WG
meetings at Oslo. Please send comments back asap.

regards,
John and Ed

=============

Minutes, day 1, recorded by John Strassner

Charter Changes and Update - John Strassner
----------------------------------------------

John described the changes to the charter. They are summarized below:
  - Information model and LDAP schema were separated into 2 drafts:
      - draft-ietf-policy-core-info-model-00.txt
      - draft-ietf-policy-core-schema-04.txt
  - Both were explicitly linked to the CIM and DEN models
  - The charter was updated to call for explicit coordination between
    the core and QoS policy information model and schemata and:
        - PIB(s) and MIB(s) developed in DiffServ
        - COPS-PR developed in RAP
        - working with IPSP
  - A document combining requirements and use cases will be added
  - A new set of milestones were added

Core Information Model and LDAP Schema Drafts - Bob Moore
------------------------------------------------------------------

The Core Information Model  and Core LDAP Schema were presented by Bob
Moore. Bob first presented the design for splitting the classes that
were in the ...-core-schema-03 document into classes that are in
...-core-info-model-00 and classes that are in ...-core-schema-04.
This also gave insight into why certain new classes that appeared in
the LDAP core schema only were necessary.There are no fewer and no
more classes than there were before, they are just organized better
and have had more detail added.

The information model specifies the semantics of policy independently
of any implementation (including directories that use LDAP as their
access protocol). It does this by using generalized object classes,
and associations and aggregations to relate the different object
classes to each other. It is an enhancement of DEN. Its purpose is to
enable other organizations to map a single model to multiple storage
technologies, making cross-technology interoperability possible.

The purpose of the LDAP core schema is to map the data and
relationships represented in the core information model to a form that
is implementable in a directory that uses LDAP as its access protocol.
Therefore, the basic semantics of these classes are referenced, not
repeated. Modeling constructs, such as associations and aggregations,
are usually represented as DN pointers and "structured DN pointers"
(this latter is a DN pointer that has semantic meaning represented as
attributes). Sometimes, auxiliary classes are also used. Specific
optimizations to simplify identification of classes in the DIT that
should be under policy control, DIT navigation and bulk retrieval of
policy-related information were introduced. These are all required to
help mitigate performance problems when using LDAP, and therefore have
no corresponding classes in the Core Information Model.

Ryan Moats offered to help describe the format of both the core
information model and core LDAP schema classes in formal ASN1. In
addition, would like to examine the concept of versioning.

Going to last call was discussed. The basic problem is that the core
information model (and to a lesser extent, the core LDAP schema
mapping) is needed by other working groups. This argues for moving
these documents to last call, once changes have been incorporated (see
end of the minutes for more info on this). On the other hand, some
people expressed concern that related work, such as the architecture
document and the PIB(s), should be progressed first, and then have the
core documents moved to last call.

One possibility is to move the document through working group Last
Call to ensure that there is consensus, but then wait to submit it to
the IESG until the architecture document (and maybe others) have had a
round of comments. The chairs and the ADs will work out a strategy and
propose this to the working group.

Raju Rajan - QoS Schema
---------------------------

Raju gave an overview of the draft that he has co-authored. It
describes a schema for modeling provisioned and signaled QoS. The
goals of the provisioned QoS section are:

  -  Identify flow aggregates
  -  Ensure resource allocation for flow aggregates using PHBs
  -  Ensure support for AF and EF
  -  Support packet (re)marking

The goals of the Signaled QoS section are:
  -  Accept or reject reservation requests
  -  Per-flow, per-user, per-domain, per-application
  -  Assign pre-emption priorities
  -  Support RSVP policy objects

The draft attempts to support range of client types, including end
hosts/servers as well as networking devices such as routers and
firewalls. For each device category, there are different actions that
are applicable. These include the following:

  Applications and Hosts
    -  User and application needs
    -  Packet marking

Edge routers and firewalls
    -  Traffic verification
    -  Shaping
    -  Aggregation
    -  Resource allocation

Backbone routers
    -  Aggregation
    -  Resource allocation

The DiffServ actions are provided through three classes -
diffServAction (subclass of policyAction), and diffServResourceGroup
and RSVPResourceGroup, the latter two of which are both subclasses of
Top. Traffic profile descriptors and packet marking descriptors are
embedded in the diffServAction class. The traffic profile descriptor
is used to classify traffic as being in- or out-of-profile. The packet
marking descriptor provides the ability to mark the DS byte for
enabling simpler QoS classification at downstream network devices.
The resource group descriptor describes the treatment expected by
packets within a common service group. The descriptor does NOT aim to
describe how the service is to be provided, as this is
device-specific. Instead, it describes the semantics of the traffic
conditioning that is to be performed.

The class RSVPAction, which is a subclass of policyAction, contains
policies to be applied to RSVP PATH messages and RESV messages. The
draft supports simple forms of policy based control, enabling
administrative restrictions to be placed on the amount of resources
that a single RSVP flow, or group of flows, may consume. Support is
also provided for an RSVP reservation to be mapped into a DiffServ
service category.

Raju presented three approaches to the following problem: if 3
descriptors are dumped into a single action class, then if there are
two domains that require separate policing that share a common
resource, this can't be accommodated. His recommended solution was to
combine traffic descriptor and marking in two different diffserv
action classes, and use a pointer to a resource descriptor. You can
further expand on this by cascading diffserv action classes, that each
have their own mark and resource descriptor objects. Problem is that
there are too many objects (Ed note: if the objects are reusable, then
their quantity doesn't matter).

There were a number of questions on the draft.

Shai: No higher-level above RSVP or DS
Raju: take rate. It's defined differently in ATM, DS, etc.

Raj: we need a separate conceptual model

Shai: Why is there no such thing as Gold Service. How do I map between
AF11 and AF31?
Raju: embedded in the condition, and the PDP does this.

Shai: See a higher-level being mapped into RSVP and DS. This is too
low-level.

Raju: What about concurrent work going on in the DMTF? Those of us
that aren't members can't see this.
John: The latest version of CIM is publically accessible

Andrew: needs alignment with terminology draft
John: agreed, but we need to reconcile this draft with other QoS
drafts to be presented
and then align the terminology

David: conditions draft enables you to build an mf classifier; what is
lacking is network state
Tos marking needs alignment with DiffServ. AutoStart should be
optional

Yoram Bernet/Andrew: is this draft at the right level? Similar to the
conceptual model
in diffserv, but why is the draft so close to the device

QoS Device Information Model - Walter Weiss/John Strassner
----------------------------------------------------------------

Best Effort was a Good Thing. Everything was straightforward and easy.
You pushed out the bandwidth, and whatever happened happened.
Congestion wasn't bad - it told you how much you had to cut back, and
how much you had to add.

Then came QoS. Things got a LOT more complicated, especially
management. This took the form of increasing complexity required to
manage individual devices to support network services.

What is a QoS Service? There are two different elements of this
service configuration and management. Policy comes more into play in
the maintenance of a service. Policies apply to services, and also
help implement services.

Need for low-level policies. Device has a set of capabilities, and we
can talk about sets of devices that share sets of capabilities. This
is especially true of DS, when we are talking about behaviors and how
services like AF and EF are managed and maintained. We're advocating
support attributes and classes in the schema that enable us to manage
such services, and to facilitate the mapping of information in
schemata to MIBs, PIBs, etc.

Slide # 6 shows the inherent sub-ordinate relationship of policies.
This is why examples such as "Gold Service" are fraught with trouble,
since "Gold Service" consists of a set of sub-services that, together,
provide the higher-level service referred to as "Gold".

There are 2 components of a policy: grammatical semantics and
attributes. Is User = George, or is this IP Address part of the Sales
Force.  Variables and constants represent the things that you want to
manipulate and check (e.g., device and service state). When you
combine these 2 concepts, you can construct policies to manage
services and business objectives. This is why low-level policies are
so important - Gold Service is meaningless unless it can be related to
what the device is capable of providing.

Shai: 2 statements in fitrst group - is this procedural.
John: No, this defines the enabling of a new class. The point is that
some policies may need to be executed in a device, whereas others
should be executed in a PDP. It depends on the component of the policy
(e.g., are you testing a group of counters)

Attributes are stored in classes, and the policy core model defines
the class structure and grammatical semantics to do this. However, we
worry about attributes being subclassed out of the policy core schema.
Problem is that if all policy attributes are subclassed from the core
schema, this leads to massive duplication of attributes, not just
between different ietf wgs, but also between dmtf models and ietf wgs.
This draft makes a clear differentiation between state and
configuration attributes, something that has been heretofore lacking.
There are really two different types of attributes - state vs.
configuration attributes. State attributes are attributes that
represent the currently active state of the device or service, eg.,
counters. Config attributes specify what the desired value that you
want the device or service to have. E.g., if average queue depth was
10K bytes, and if we want to change that for a certain set of devices,
then we can decouple state from configuration. The point is that we
shouldn't be shackled by the protocol. You can also represent a group
of devices, or interfaces, or services, or other objects.

Contexts and roles. Roles need to be formally added to the grammatical
syntax of policies. These are used to define what policies are applied
to what devices.

Raju: Why can't we use the policyGroup container for roles?
John: Because a role is simply a string attribute that defines the
name of the
object in question. It is not a container. Furthermore, a container
has the
limited semantics of being able to hold objects. A role points to an
object
that has much more powerful semantics, describing how it is used in
the system.

Question: Where does this fit in w.r.t. the bigger picture?
Walter: We want people to think about policies in respect to low-level
and
high-level function. We also want to differentiate between
configuration
and state.

Question: how does this scale? Suppose you had a relatively large
network.
How many objects would be required?
John: Impossible to answer, since policies are inherently different in
nature,
and it depends on what it is that you are trying to do. For example,
controlling a single device will generally take less objects than
controlling a
set of devices. But, as we'll see in the next presentation, being able
to reuse
objects mitigates their sheer number.
Question: But isn't this a lot of objects?
John: Again, impossible to define in the general sense. But it is
generally
greater that 1 and less than 50.

===========
Day 2, minutes recorded by John Strassner
===========

Yoram Snir - QoS Information Model
---------------------------------------

This draft is complementary to the previous draft presented. This
draft subclasses the policy core information model, whereas the
previous draft subclasses the network model of CIM.

This draft is an extension to the core shcema draft, in spirit and
structure. Aimed to help interoperability. Schema is made up of 3
parts: Scoping, Repository, and Definitions.

Reusable-objects. This is a notion currently missing from the core
schema. We feel that it is essential to building highly scalable and
reusable schemata. The repository classes are subclassed from the core
schema. This enables us to add semantics to the container classes
defined in the core schema.

The repository structure is described. Note that it is extensible by
nature, what is presented in the draft is meant to be a foundation.

Reusable objects are placed in the repository. This lets you abstract
key concepts (e.g., an HTTP port can be referred to by HTTP, and you
don't care what the exact port number is). It also helps protect
against some of the deficiencies of LDAP - if classes reference
reusable objects from one repository class, then changes can be
simplified - always get them from one place.

An extensible containment hierarchy is defined in the draft. This
divides objects into multiple domains, each of which can have a set of
named containers that contain groups of policyRules to be applied for
a common set of applications, or to a subnet, or to the employees of
an organizationalUnit, etc.

Why qosPolicyRule? Because the policyRule class is generic and nature,
and certain key attributes needed for QoS are missing. One example is
priority, which is essential if multiple PDPs need to get to the same
target object.

Simple conditions are built out of variables and constants. Variables
have a name and a type, and represent constraints. Constants have a
name and a type, and represent values. We've defined some simple value
classes; more can be added as required.

With this structure, we are able to model provisioned and signaled QoS
conditions and actions in an extensible way. We also have the ability
to model simple and complex conditions, as defined in the core schema
draft.

Something that we're working on is how to abstract the PHB. We're also
working on general definition for PATH and RESV admission control.

Simple condition example - attachment
This shows a simple condition built out of the attachment of a
variable and a constant class. Note that the constant class has a
qosPolicyNumberValue class attached to it. This is in the spirit of
the core schema, and enables you to retrieve this in a single access

Simple condition example - reference.
Gives you more flexibility at the expense of more accesses. But more
accesses aren't bad if the objects that they are referring to are
reusable.

Complex condition
This is also accommodated by our approach. Consists of a set of
references to separate objects

Simple QoS Policy Rule
Example consists of a SimpleCondition and a single action. Each of
these can be attached directly to the qosPolicyRule, which is a
structural class.

QoS Policy Rule example
Now we have a qosPolicyRule with a condition List and an actionList.
Each of the lists point to other objects. The objects themselves can
be simple objects or objects that have other objects attached to them.

Namespace example
This reflects the translation of different administrative domains.
Within a given domain you have a set of policy rules that can be
simple or complex.

David Black: good that someone tried to put this together, likes it in
general

Glenn Waters - Status of the Policy Framework draft
------------------------------------------------------

There is a new set of authors (Hugh Mahon, Mark Stevens, John
Strassner, Lee Rafalow, and Glenn Waters). Next revision to be
submitted to the draft editor by 31 August, will have an interim copy
available for wg review by 15 August.

The general architecture was described. Focus was on the partitioning
of the framework into four components (Policy Management Tool, Policy
Repository, PDP, and PEP), and their interconnection.

The purpose of the policy management tool is to serve as an interface
between policy definition, editing, and management functions external
to the Policy Server (e.g., the repository, PDP, and PEP) and the
components of the Policy Server. It consists at least of an editor, a
means to translate policy rules entered in human-readable form to a
format compatible with the type of the repository being used, and the
ability to validate policy rules. This includes syntax and semantics
(to as much an extent as possible) checking, as well as simple static
conflict detection.

The repository is used to store policy information and rules. Note
that any technology can be used, but particular attention is paid to a
directory that uses LDAP as its access protocol.

The PDP performs three main functions. First, it responds to policy
events and locates the appropriate rules that are of interest to this
event. Second, it performs state and resource validation (for example,
bandwidth broker using policies to control bandwidth allocation).
Finally, it converts the policy rules as stored in the repository to
device-specific mechanisms.

The PEP is responsible for executing the appropriate actions. Note
that the PEP may in fact evaluate the policy instead of the PDP, if
the conditions of the policy are (for example) traffic conditions that
PDP doesn't understand.

Other important parts of the document discuss conflict detection and
interoperability (e.g., how the data in the repository is used and
accessed).
Talk more about Cisco additions, but note that there is no existing
standard to pass notifications. Note also that we don't talk about
protocols in this wg, and so we can specify its functional
requirements in the architecture, and we can specify what happens, but
we can't specify the protocol itself.

Document sections left to write:
  - Roles: John
  - PIBs:  Hugh/Mark
  - Conflict detection additions:  John
  - Relationship to other WGs:  Hugh
  - Procedural vs. declarative:  Glenn

Keith: clarify that you will have multiple of each component (e.g.,
the PDP)
Glenn: OK

Fran: the device adaptor never disappears
Glenn meant a translation of format.
Fran: Doesn't this change how things are stored in the repository?
Glenn: No, it just means that the PEP is very smart and knows how to
deal with information in a raw format

Yoram Bernet: Need to beef up sections on RSVP (e.g., COPS for RSVP)
and generalize things like filters for both DiffServ and RSVP usage

Keith/Nitsan: worried about distribution of policy information

Raj: not trying to solve LDAP problems, right?
Glenn: right

Yoram: use of directory - just for static info, right?
Glenn: right

Walter: variable manipulation can be part of the info model, but do
NOT have to be pushed up all the way to the directory

John Strassner - Update on the Terminology draft
--------------------------------------------------

The purpose of this draft is to define a standard set of terms that
describe policy. It should be compatible with formal modeling
terminology.

Motivation was to reconcile differences in terminology between the
DiffServ, RAP, and Policy groups. Intent is not to redefine
terminology of other groups (e.g., what is a PDP) but rather to help
drive a common agreement between the working groups.

This was recently reissued as a working group draft. Final edits are
being incorporated, and then the next step is to do a thorough review
of it with the updated DiffServ and RAP documents. Will then reissue
by end of August.


Hugh Mahon - Policy Usage and Requirements

This will turn into a working group draft.

The focus in on how one will use a policy system. It emphasizes the
separation of policy delivery from the definition, editing, and
management of policies.
It covers issues such as:
  - accommodating those particular policies that need to be
    delivered in a timely manner. Note that this is separate
    from how often the policy changes, or who changes it.
  - providing feedback to the administrator describing what
    happened to the policy (e.g., was it delivered? were
    there any installation problems? etc.)
  - how does the admnistrator know that the authored
    policy is what was received?
  - how to deal with the lack of transactional and
    referential integrity in LDAP
  - role management

Nitsan: reporting policy server errors. This has a usability problem -
we need to be able to tell the user that a particular policy was
successfully downloaded, installed, and deployed.


Scott Hahn - Description of the PIB
------------------------------------

Important point is that PIB lives not in the schema but is passed down
to the device. Scott views schema as high-level and abstract, not at
the level that the device needs to implement policy. PIB is
responsible for configuration. It enables the same configuration to be
pushed down to a set of devices.

So, the PIB is not a protocol, nor is it a schema. It is a data
definition that exposes the capabilities of a PEP to the PDP. The PEP
uses the PIB to report device capabilities to its PDP. The PDP uses
the PIB to provide configuration information and updates to the PEP.

Policy server and PDP developers need to be aware of both the PIB and
the schema, because they need to map information contained in one to
the other, and vice-versa. Referring to slide #3, PIB information is
carried on the wire between the PDP and a PEP.

The flow of data is simple - the device expresses its capabilities to
the PDP, and then the PDP matches policy rules to those capabilities
and sends down configuration information. The PEP can then report
status and/or errors to the PDP. Note that the PDP can asynchronously
update policies as required.

The PIB is protocol-agnostic, but one specific implementation was
written with COPS-PR in mind as the transport protocol. The PIB could
easily be carried over SNMP, for example.

The initial draft focuses on a QoS PIB, though the concept is
extensible to other policy domains. Now part of the DiffServ WG.
Focused on DiffServ policies. QoS PIB consists of 3 modules: general
info, IP module, and 802.1p module.

The PIB uses the concept of roles and role-combinations.

More integration needed with DiffServ conceptual model and the
DiffServ MIB. Metering currently is not addressed.

Kathy Dally - How is violation of Policy handled?
(we need to take this to the working group)

Walter - we need a practical example of how to use a PIB. In
particular, we need to substantiate the claim that it contains
lower-level information than the schema contains.
(again, we need to take this to the working group)

Policy Conditions (Raju)

PolicyGroup groups policyRules. PolicyRules talks about
policyConditions and policyActions.

What shoul be the content of policyConditions, and how should it be
structured. PolicyConditions need packet header and PEP information,
administrative categories, and network state/event categories. This
should be represented in the schema, but we must consider data and
query costs, extensibility, structural vs. aux classes. These
considerations affect the overall object model.

Administrative categories - user, user groups; hosts and host groups;
apps; roles

Network state/event information. Wide range of events that are
relevant to QoS. But you get into trouble when you try to put this
into the directory. Examples of this are network congestion, routing
information, or RSVP path message arrival. How will we achieve
interoperability, and how can we ensure that multiple devices from
multiple vendors will use the same definitions.

Policy condition - if you embed an expression directly as an attribute
in the condition object, then it isn't extensible and you need another
schema to interpret it.

One way to do this is with one structural class with all may
attributes. Not extensible, and doesn't allow one user to extend
attributes of the class of one type without affecting the entire
class. Preferred way is to use multiple aux classes.

Jamie Jason - Working with IPSP
---------------------------------

There are currently 3 drafts in IPSP - ASN1 draft for VPNs, another,
plus UML draft. We've unified the 3 drafts in order to design policy
to control IPSEC in the abstract. After we get agreement from the
IPSEC wg, we want to get agreement with the PFWG. This will enable us
to ensure that their draft is properly derived from PF core classes,
as well as enable them to provide direct feedback to this wg. Want to
be able to represent the 5-tuple somewhere, maybe not in core, but
certainly not repeatedly in IPSP and DiffServ.

Final Action Items (Ed Ellesson)
--------------------------------

General consensus that some concepts in the Snir-Ramberg-Strassner QoS
draft (e.g., repositories) need to be moved into the core policy
document. Authors of both documents to get together and plan this out.
Then, both documents will be re-issued.

Policy terminology draft to be updated, and a new revision sent to the
list. It is hoped that the next revision of this (and the policy core
information model draft) will be ready for working group last call.

Core LDAP draft proceeding well, but needs another revision before we
can start considering Last Call.

The authors of the framework draft will produce an updated draft on or
before 31 August. Hugh will try to produce his new draft by 31 August
too.

It seems like we should have an interim meeting, possibly one day of
design team from Policy, DiffServ and RAP, and another day of the
Policy Framework working group. Target date will be early September.



