Chairs opened the meeting with agenda bashing and document status update (documents accepted as WG documents): draft-ietf-dnssd-hybrid-01 * Will consider for WG last call today * Does this document meet the requirements of RFC 7558? * Looking for document shepherd draft-ietf-dnssd-push-02 * Needs review against dnssd-hybrid specification * Will consider for WG last call today * Looking for document shepherd draft-ietf-dnssd-mdns-dns-interop-02 * WG last call complete 2015/07; comments discussed at IETF-93 * Rev -02 published includes updates from WG last call comments * Needs final review by WG before forwarding to IESG * Looking for document shepherd The chairs asked if there was consensus in the room to do a short WG last call review to confirm that draft-ietf-dnssd-mdns-dns-interop-02 addresses the comments during the previous WG last call. Summary meeting results: ------------------------ draft-ietf-dnssd-push: after 2nd implementation ready for WG last call draft-ietf-dnssd-hybrid: ready for WG last call; will send to IESG when draft-ietf-dnssd-push is ready for IESG draft-ietf-dnssd-mdns-dns-interop: ready for WG last call review and then send to IESG Document discussions: --------------------- ------------------------------------------ Scalable DNS-SD (SSD) Threat, draft-otis-dnssd-scalable-dns-sd-threats-0 ------------------------------------------ Doug Otis gave a presentation about the -02 revision, which replaces draft-otis-dnssd-mdns-xlink-06 anddraft-rafiee-dnssd-mdns-threatmodel-03. Based on WG comments, draft-otis-dnssd-scalable-dns-sd-threats-02 needs additional editing to use a more accurate model of DNS, mDNS and other name resolution mechanisms as the basis for the threat analysis. Doug Otis(DO) presenting: Andrew Sullivan (AS): Talking about DNS and not talking about DNS. Problem is don't have a consistent namespace, then there is mDNS, then Public DNS Namespace. Threats identified consequence of multiple name spaces. DO: Agree, but how do you express this in the draft. AS: Private DNS is an alternate root. DO: May be in Private Root, may be in Global Root AS: Only one view of that Global Root DO: Different views? AS: External View and Internal View AS: This is not a terminology dispute. Paul Hoffman (PH): Theme is security issues. They only come up when you mis-understand there are two name spaces. Pushed together two different separate namespaces. Ron Atkinson (RA): Agree with AS and PH, content to document. obfuscated but not intentional. Document needs a rewrite. AS: IAB said years back there is a Single Namespace of the Internet. DO: Agree with what has been said and understand concerns. Problem is how to explain clearly enough for folks to understand. AS: It is okay to say "This is Bad" because "This is Bad" Ralph Droms (RD): Take to the mailing List. ------------------------------------------------ DNS Push Notifications, draft-ietf-dnssd-push-02 ------------------------------------------------ Stuart Cheshire gave a brief review of updates to the -02 rev. The WG then discussed whether the document is ready for WG last call. Consensus is that the WG would like to see an interoperable implementation prior to WG last call. Stewart Cheshire (SC) presenting: Added updated based on comments in last meeting Feel ready for WGLC Ray Bellis (RB): edns-keepalive has been redone, should be revisited. DO: if queries are multicast, build ptr record sets SC: as a debugging tool, one can ask for service types; then ask for instance of a service type DO: overlay hostname is a service. ----------------------------------------------------------------- Hybrid Unicast/Multicast DNS-Based SD, draft-ietf-dnssd-hybrid-01 ----------------------------------------------------------------- Stuart Cheshire gave a brief review of updates to the -01 rev. The WG consensus in the room was that the document is ready for WG last call. However, because this document depends on draft-ietf-dnssd-push-02, this document will be held and then sent to the IESG when draft-ietf-dnssd-push-02 is ready to go to the IESG. Stuart Cheshire presenting: AS: Don't care if you use this, Dyn says no one pays attention to less than 30 seconds. SC: More relevant in the Enterprise Dave Thaler (DT): Similar exercise with home gateways, which caches for no less than 30 seconds. Markus Stenberg (MS): timetable for deployments AS: Values in document of 10 seconds, etc. Time give with push, any chance of experiments to get the best values for this? Also, think this is ready for WGLC SC: Made it specific on purpose rathar than vague. DT: Positive number can be higer SC: this could be any renumbering event. MS: Prefer smaller than larger numbers Optimizing DNS-SD query using TXT records, draft-aggarwal-dnssd-optimize-query-00 Dave Thaler reviewed the previous presentation of draft-aggarwal-dnssd-optimize-query-00 and presented additional details of allseenalliance.org/framework/documentation/learn/core/system-description/advertisement-discovery. The AllSeen Alliance document describes extensions to DNS-SD/mDNS defined by AllSeen. The request from the WG is to write a document accommodating AllSeen requirements while adhering as closely as possible to draft-ietf-dnssd-hybrid-01. Dave Thaler presenting: SC: Thanks for the work. No knowledge of any of this. Interesting that Israeli routers only pass Bonjour packets; usually the reverse is true (in response to Dave Thaler's motivation in the presentation for AllJoyn's use of the mDNS port).. Unicast routers partially supported. RD: Optimizations in wifi which will play badly with this. SC: Different Devices/Attributes can't be described in a single bit. Samita Chakcrabarhi: consider other things than wifi like 6lo, then DNS discovery needed DT: Is this an interesting way of solve this problem? SC: Will work on something with collaboration AS: On the filters (third one), appetite for doing more extensible way. Attractive Nuisance. Chairs wrapup: RD: WGLC on next rev of hybrid proxy; that will then wait on PUSH, which Stuart will revise, WGLC when we have interoperable implementations WG will discuss what to do with concepts in the doc Dave presented.