IETF 59 (Seoul) OPS Area Open Meeting - Minutes Tuesday, March 2, 2004, 15:45-16:45 Slides at http://psg.com/~bwijnen/OpsArea59.ppt Minutes based on notes by David Harrington. Agenda: 5 minutes agenda bashing/admin/assign minute-takers 1. 5 minutes New AD intro 2. 10 minutes Division and Status of WGs. 3. 5 minutes Quick overview/status of OPS directorate and MIB Doctors 4. 15 minutes Other submitted or requested presentations (OPS related of course) 5. 20 minutes Open Mike and Q&A 1. Introduction to new coAD David Kessens. No questions. 2. Review of WGs and division between ADs. 2.1 WGs for which Bert Wijnen is primary AD: - Bridgemib - discussion of moving further MIB devevelopment in the 802.1 space to IEEE 802.1, with MIB doctor review in IETF. This was also discussed during (informal) liaison discussion at the IEEE 802 meetings in Vancouver in January. We still have a few bdirdgemib documents though that need to be finalized. We need volunteers. - Capwap - Initial doc submissions for this work reviewed. It did not look right for IETF to define functional blocks for access points. So we have been talking to IEEE 802.11 to get them to define the architecture. The capwap WG will gather a taxonomy and current practices, and possibly recommendations, as input to potential IEEE 802.11 work. When IEEE is done, we can then work on management and provisioning protocols. - Disman - work is close to complete. Chair may drop some items because of lack of volunteers/interest. All specific milestones have been completed, and there is no groundswell for new work. Chair recommends closing the WG. - Entmib - was close to done last time; working on their last set of docs. - Hubmib - has completed all current milestones. Have new work for EFM (Ethernet in the First Mile). Working close with IEEE 802.3. - Imss - IPv6 over fiber channel; in IESG hands; will go out for IETF last call soon. MIB work will start later this year. - Ipcdn - cable modem mib documents; aggressive dates are not being met. - Netconf - chairs believe they are closing issues, and protocol will be able to finish soon. Basically the protocol got a "haircut" as Elliot expressed it later in the Netconf WG meeting. Data modeling - people have met (informally) to start discussing this work. A separate mailing list has been set up for that, since it is not a WG charter item at this time. - Policy framework - last document has been through RFC AUTH48 review. So RFC is out or should show up real soon now. One individual document for LDAP Schema for PCIM extensions has been discussed on list; not a WG work item though. WG will be closed soon. - Rap - almost no work over last year; last document (cops over TLS) is going back and forth between AD/IESG and editors. Hopefully we can close soon. - Rmonminb - ongoing. Raqmon overdue, and serious issues were raised by the AVT chairs. Will need revisions in the next few months. - Snmpconf - one document left in IESG discussion. Once approved (hopefully soon) then WG will be closed. Questions: Randy presuhn: to what extent is it worth expending energy on advancement from PS to DS if we expect changes to process (newtrk WG)? Bert: If docs for adavancement are close to completion, then complete them. If you still have to start the work and there is no urgent need to advance, then you might want to work on implementation reports, so that you can use that when/if new stds track levels get defined. Avi Leor: in Minneapolis, at radius extentions (RADEXT) BOF, the vast majority wanted a WG. Why isn't there a WG or a BOF at this meeting? Bert: Randy Bush was the AD at the BOF, so I didn't have data on hands raised. BOF Chairs and list reached consensus that existing radius should not be changed and that extensions (if any) should be compatible with existing RADIUS specs. Extensions need to be defined, but justification is required. AAA/Diameter is meant to take over, so it will be hard to justify some of the work in RADEXT. BOF chairs are trying to line up volunteers. Avi: How does not having a BOF generate interest? Bert: work can start prior to a WG starting. Individual drafts have been submitted already. The drafts mmust be within the charter limits for backward compatibility, etc. We are talking with other organizations that need these extensions. Both BOF chairs and I agree it is not yet time. Joel Halpern: there is disagreement to what consistitutes backwards compatibility. Most vendors do not want Diameter; lots want radius extensions. Bert: we don't just want to open a WG that could be a multi-year effort with extensions that don't belong in this space, so we need to be careful. Farid: I am part of GSM Association (GSMA), and we sent an official liaison to the chairs. About some of the drafts. Bert: yes, BOF chairs copied it to me and I am verifying with GSMA that this is an official statement from GSMA and why it is needed. I need to verify this first that authors are in fact members of GSMA. I hope you agree that this is due process. Farid: I know this Avi: there is a bar BOF to discuss this. Dan Romascanu: my impression is that there was a strong consensus that the work should be chartered. It looks like the people who were going to do the work didn't get organized very well. Everybody was surpoised when the WG didn't show up on the IETF agenda. Bert: BOF chairs have not submitted a charter to the ADs. Avi: chairs didn't even tell us the chairs were not going to be present and we could have. Bert: the fact that chairs are not at a specific meeting is not a blocking factor for having a meeting. But the BOF chairs did not feel a session would make sense. Randy Presuhn: while we're on the topic of BOF getting to WG. Another BOF was held about getting integrated security for SNMPv3. Help and inut is greatly appreciated. Dan Romascanu: IEEE WGs. I think the mode that worked 10 years ago isn't working any longer. I think we are without any plan, and bridgemib cannot be sustained in limbo. I think we need to go back to IEEE and tell them they need to take editing responsibility, with mib doctor review. This should happen with the bridgemib and hubmib. Bert: agreed. We are working on this. I support doing this for both bridgemib and hubmib. 2.2 WGs for which David Kessens is primary AD: - BMWG: no comment - DNSOP: grossly outdated charter; working to get a new charter. Agreement between chairs and Ads. - GROW: one co-chair (Vijay) stepped down; found Geoff Huston will co-chair WG. Working on charter, and have chair/ad agreement. - IPFIX: no comment - Multi6: could have huge implications for the internet. Working on charter, with lots of attention from IESG. There will be discussions of architectural implications. - Psamp: doing good work, but will be delayed due to ipfix dependence. - V6ops: very long list of work items. I'm trying to sort things out. Working with Bert on this WG. - RADEXT BOF: we will be following it closely. - Ptomaine: one draft waiting for publication, then will be closed. Questions Avi: bert said radius is dead. Bert: we should not keep trying to extend radius. We should be working to move to the use of Diameter. Avi: that is not true. Radius is alive and well, and it is growing in acceptance. Unfortunately many new areas are adopting radius. My company sells RADIUS and Diameter, but nobody has requested Diameter; everybody requests RADIUS. It appears the IESG is trying to kill RADIUS and push Diameter, despite market pressures. Bert: my understanding is the biggest concern for radius is security. SNMPv1 is also insecure, and we won't support further development on SNMPv1. Avi: sometimes security is good enough security. Elliot lear: we have finite resources, and we can work on schlock or we can work on good engineering. As much as Diameter fixes the problems of RADIUS, we aren't we pushing it even more. Avi: we should do both. If IETF doesn't extend RADIUS then proprietary extensons will be done. Your Diameter will want the extensions, and the ietf won't be able to do them. Elliot: you missed my key point. We have limited resources and cannot pile hack on hack to keep radius alive, we won't do good engineering. Avi: spending effort on Diameter, which nobody wants, is a wate of resources. The IETF has to be cognizant of market drivers as well as technologies. Bert: feel free to discuss this on the RADEXT list. Further discussion here will not be fruitfull/worthwhile. 3. Directorates: OPS directorate: helps review documents that are on IESG agenda and specifically review for Operational aspects. http://www.ops.ietf.org/ops-dir.html IPv6 directorate: new to advise on IPv6 related issues. This directorate advises INT Area and OPS Area ADs. http://www.ietf.org/u/ietfchair/ipv6-dir.html MIB doctors: review all MIB documents in the IETF, looking for good MIB syntax and structure and for reuse. Also sometimes help design MIB modules. MIB Doctors do not (always) know the technology being managed, however, so each specific WG needs to deterimine the attributes that are needed and that WG also needs to make sure that the MIB does represent the correct management information for the technology being modeled. MIB Doctors use guidelines to get consistent reviews. Editors and WG chairs should also review the document against those MIB review guidelines themselves. See http://www.ietf.org/internet-dra fts/draft-ietf-ops-mib-review-guidelines-02.txt 4. Prepared presentations: none were submitted or requested. 5. Open Mike and Q&A: |