WG Meeting Summary

The IPCDN WG met to discuss the progress of various MIB documents, especially the DOCSIS and PacketCable/IPCablecom submissions. 13 I-D updates were submitted in between the meetings. We briefly covered the rechartering of the WG under the Ops area in August and the updated milestones.

On the DOCSIS MIB front, we went over the open & closed issues on the main drafts including RF MIB v2 draft 08, DOCSIS Cable Device MIB v2 draft 6, Subscriber Mgmt MIB draft 13, DOCSIS BPI+ MIB draft 12 and DOCSIS QoS MIB draft 9. On the DOCSIS QoS MIB, we discussed the issue of the DOCSIS use of TOS byte vs. DSCP with Randy Presuhn and Bert Wijnen during the meeting and agreed on a proposed resolution path which is to complement the limited statement of applicability we already have by adding a mapping and send the text to Bert.

We also reviewed the status of the PacketCable/IPCablecom MIB Updates (MTA MIB draft 2, Signaling MIB draft 2, Management Event MIB draft 2) and the next steps (MIB doctor review).

Finally, we went over the comment Randy Presuhn had made re: the ping MIB and the CableHome Remote Diagnostic Tools MIB and agreed on the next step which is for CableLabs to proposed a compliance statement against the ping mib in order to reuse that mib at a maximum.

Total meeting time was one hour.

There were a number of draft updates (13) for the WG since the last IETF meeting in Vienna, but there have been no updates for the CableHome MIBs.

The WG discussed the co-chairs' proposal for making realistic WG deadlines for the existing drafts. This requires more input from the WG.

Due to the expiration of the draft, and after discussions with the current authors, the co-chairs have concluded that the DOCSIS Event Notification MIB needs new author(s). WG participants asked if there were outstanding comments on the Event Notification MIB; there have been some comments from some reviewers that need to be addressed.

David Raftus, who introduced most of the DOCSIS 2.0 related content in this internet-draft and produced seven revisions, has voluntarily relinquished his draft editing duties. Eduardo Cardono (CableLabs) has agreed to take over this draft, and prepare it for IESG submission. Eduardo submitted version 8 of the draft prior to the Minneapolis meeting, resolving a number of WG open issues with the draft. See the IPCDN WG presentation slides for more details.

This internet-draft is considered high priority by the WG with regards to IESG approval, since many other IPCDN internet-drafts make reference to it. A draft version 9 is already underway, and will address the three technical issues (ifAdminStatus, interface cloning process, and Security Considerations update) raised in the meeting. See the IPCDN WG presentation slides for more details.

Richard Woundy included a number of changes in version 6 of the internet-draft, particularly in adopting the DiffServ MIB for IP filtering, and subsequently deprecating the IPv4-only and IP ToS filtering tables and objects. Other editorial changes were also made to the draft.

A WG participant asked if implementors could choose between implementing the DiffServ MIB objects or the legacy RFC 2669 IPv4/ToS objects. Bert Wijnen explained that "deprecation" means that new implementations must implement the DiffServ MIB objects -- although new implementations may implement the legacy deprecated objects as well, for backward compatibility.

The WG discussed the impact of changing the upper bound on the size of docsDevSwFilename. The current size restriction is 64 octets, but there is a desire to increase the maximum size to 128 octets (or more). A WG participant pointed out that a proper current SNMP agent implementation must perform a size check for the Set operation, so there should be no adverse impacts to current implementations.

The direction of the WG is to increase the upper bound on the size of docsDevSwFilename, without defining a new MIB object. The change in syntax to docsDevSwFilename should be reflected in the compliance statements for RFC 2669 in this MIB.

Wilson Sawyer incorporated minor changes in version 13 of the internet-draft, mostly in response to MIB doctor reviews by Bert Wijnen and Harrie Hazewinkel. See the IPCDN WG presentation slides for more details. The WG agrees that this draft is ready for submission to the IESG.

Eduardo Cardona submitted version 12 of the internet-draft, in particular to address issues raised in the Area Director review of the draft. See the IPCDN WG presentation slides for more details. The WG agrees that this draft is ready for submission to the IESG.

Michael Patrick and William Murwin submitted version 9 of the internet-draft, resolving WG open issues with the draft. See the IPCDN WG presentation slides for more details.

The remaining open issue with this draft has been its reference to the IP TOS octet, rather than to the IP Differentiated Services field (DSfield). The rationale for this approach is that the DOCSIS specifications specifically refer to the IP ToS octet rather than the DSfield with respect to the objects in this MIB, and that it is critical for this MIB to model the DOCSIS device state accurately.

The objects in question include: docsQosPktClassIpTosLow, docsQosPktClassIpTosHigh, docsQosPktClassIpTosMask, docsQosParamSetTosAndMask, and docsQosParamSetTosOrMask. Bert observed that all of these MIB objects are read-only. Randy Presuhn made the comment that these objects should not pose a problem as long as the conversion from the IP TOS octet to a DSCP value is well documented in the internet-draft. He observed that this level of documentation is not in the draft yet and probably should be added.

The WG agreed on the following resolution to the IP TOS octet concerns.

Note: After the WG meeting, Richard Woundy observed that docsQosServiceClassTosAndMask and docsQosServiceClassTosOrMask are IP TOS-related MIB objects that are read-create objects in the MIB. The WG needs to confirm that the Area Director guidance above is still applicable for this draft.

There were no comments on the other QoS MIB changes for version 9. The WG agrees that after addressing the IP TOS-related issues above, then this draft is ready for WG last-call, and then submission to the IESG.

Eugene Nechamkin and Jean-Francois Mule incorporated changes in version 2 of the internet-draft, to address comments from an initial MIB doctor review, and from the WG last-call.

In the month since the draft was submitted, there have been very few comments. The first MIB doctor (Mike Heard) doesn't have time at the moment to finish the review. The co-chairs need to remind Bert Wijnen to help find a new MIB doctor to complete the review of this draft.

Gordon Beacham, Satish Kumar, and Sumanth Channabasappa submitted version 2 of the internet-draft, to address comments from the WG last-call. This draft has had no MIB doctor review yet.

ETSI and CableLabs have been collaborating closely on the NCS Signaling MIB. Jean-Francois asked about when should formal comments be solicited by ETSI and ITU -- before or after the MIB doctor review? The next ITU SG9 face-to-face meeting is in January. Bert responded that it would be preferable to get a MIB doctor review before the SG9 meeting; Jean-Francois should ping Bert prior to the meeting.

Wim De Ketelaere and Eugene Nechamkin submitted version 2 of the internet-draft, to address known comments from ETSI and the IPCDN mailing list. The recent changes included aligning the SYSLOG precedence values with the DOCSIS Cable Device MIB and the SYSLOG MIB. A previous draft labelled as version 2 was never posted to the IETF.

This draft is ready for a MIB doctor review.

There have been no new submissions of the CableHome MIB internet-drafts since the last IETF. There have been few comments, except for comments from Randy Presuhn, which are discussed below.

Randy asked why CableHome does not re-use the remote diagnostic tool work from the DISMAN WG. In particular, he asked whether a compliance statement could be developed to include objects from the Ping MIB, rather than defining a new MIB.

The DISMAN WG is looking to promote the Ping MIB from Proposed Standard to Draft Standard status. This might permit an opportunity for DISMAN to specify new Ping MIB object or conformance groups, that make it easier to leverage in external MIBs, such as the CableHome MIBs.

Separately, CableLabs has analyzed the Ping MIB; the analysis has discovered an 80% overlap (for similar objects). The CableHome Tools MIB only permits one test at a time, whereas the Ping MIB permits multiple tests at a time.

The CableHome MIB authors have agreed to work on incorporating Ping MIB compliance statements for feedback from the DISMAN WG soon.

The co-chairs believed that the DOCSIS RFIv2 MIB and the Cable Device MIB can be finished by December, if MIB doctors have the time to review (and if the authors cooperate through submissions and responses).

The co-chairs will review plans for promoting, finalizing, and reviewing current IPCDN MIBs. They will look for serious commitments from authors to make sure the WG meets revised milestones.

Co-chairs: