Open OPSAREA meeting minutes 2010, March 23 ? Anaheim, CA (based on notes taken by Juergen Quittek and Sujay Gupta) --- 1. NETCONF for Smart Grid presented (Andy Bierman) Andy Bierman explained properties of the NETCONF model that may be relevant for its aqplication to smart grids. Data models support different naming authorities. Further advantages of NETCONF is the automation of configuration tasks, particularly for frequent changes or changes of many devices. It also can be used for monitoring, particularly since it supports server-side filtering. Finally, a very important feature is that the server advertises its capabilities. The roles of client and serv er are clearly defined and not interchangeable. Dan Romascanu: Is this useful for smart grids? Andy: I have not yet understood what is special about the problem. I am not recommending NETCONF be used. I just point out some advantages. Tina Tsou: Can Netconf be used for smart metering and dynamic pricing in smart grid? Andy: Yes, you can find an appropriate datamodel and you can define your own operations for specific tasks. Tina: Is there adaptive work needed besides modeling? Andy: There is a concern with using netconf sessions because of the significat overhead in the context of managing thousands of smart meters. It inclides starting TCP, SSH, and NETCONF session for each of them. unknown speaker: Where can we find out more about it? Andy: There is a web page about smart grid. Parts of NETCONF are appropriate, parts are not. Juergen Quittek: I think you did well in rather describing NETCONF features and not its applicability to smart grids. There are so many notions of smart grids around the world that we cannot make statements on smart grids in general. Only for a specific smart grid, we might be able to answer the question Andy: the reason we brought this up is that NETCONF was one of the candidates in the smartgrid discussions. . --- 2. IETF outcomes wiki pages (David Crocker) David Crocker reported on the new wiki pages set up for the OPSAREA -- sign is v bad - is not bad ++ sign is vgood + is good there is a ML for discussing the wiki. unknown speaker: It's great that we have a feedback feature. David Crocker: We shouldn't have technical discussions in the Wiki. They belong to the specific mailing lists. I want to keep the pages simple and not add too many features. Andy: Let's add a column for standards track status David Crocker: The standards track status does not matter here. The critical point here is: Does it get used? unknown speaker: It would be a good thing to have the standards track status and it's not much work adding it. --- 3. Security Considerations for YANG modules Dan Romascanu: We have boilerplate for MIB modules. Now we also need them for YANG modules. A first draft of security consideration section guideline has been sent to the NETMOD mailing list. It will be forwarded to the OPSAREA list. First point >> concern about the writeable nodes in the yang scheme Second point>> they are readable, therefore may become sensitive to list the modules which are susceptible. .... provide i/p to bert which can be disussed innetmod wg second phase to go ahead is speak with security AD's --- 4. NETCONF implementation report (Ray Atarashi) Ray described a network management system built with NETCONF for managing hosts and network equipment in a datacenter. Implemented functions include configuration and resource management. Joel Abbey: Which data modeling language did you use? Ray: We used an XML schema Joel Abbey: Is the module available? Ray: Yes. We built our own model two years ago when there was no YANG Randy Bush: Thanks for doing and implementation. They did not use a data modeling language of the IETF? David Partain: YANG did not exist when they started. Andy Bierman: The models that we have now are not perfect, but useful --- 5. WG Status reports: V6OPS: met yesterday, 15 documents under discussion, but only few WG drafts. close to closure of CPR security draft describing a firewall One work item on transition mechanisms has been moved to softwires. The draft on advanced security looks interesting to operators. RADEXT: radius extensions guidelines are still heavily discussed. More work going on on IPv6 attributes. Also how to run RADIUS over TCP is being discussed. PMOL: SIP metrics draft has been submitted to the IESG. The SIP benchmarking draft is moving toward WGLC. Still workin on the framework draft. OPSEC: 2 WGLCs since last meeting. One is in expert review and one lacks feedback. It looks like there will be new work on IP options handling. Router control plane protection may also become a new work item. OPSAWG: Some RFCs published since last meeting. One I-D at the IESG Bog topic: what does OAM mean? Several new drafts in the energy monitoring space. NETMOD: Basic YANG specification and data types are in IETF last call. Next step will be generating core YANG models, such as an interface model. Need to broaden the number of people working on this NETCONF: Basic work done for some time already, NETCONF monitoring submitted to the IESG, 4741bis and 4742bis in work see NETCONF Wiki. MBONED: no of papers on amt problems brought up by at&t, needs discussion udp checksum is a big issue multicast mobility several performance drafts presented almorton >quest: was the at&t discussion on the mbone list IPFIX: IPFIX MIB completed, IPFIX configuration using YANG almost complted Mediation is still the big issue in the WG. All remaining drafts have new versions and there are several candidates for new work items. GROW: bmp fib compression DIME: Has new chairs. New work on NAT control. Several drafts in RFC Editor queue CAPWAP: completed all WG items, MIBs still in the RFC editor queue closing expected at next IETF meeting ADSLMIB: Not meeting this time. New work item is DSL vector MIB 6. Open mic Tina>> suggest workshop for diameter and radius , goal is to encourage more people to be engaged in the review of the diameter and radius ongoing drafts and provide their feedback from their operational perspective