IETF-102 ippm agenda Session 2018-07-18 1330-1500: Saint-Paul/Sainte-Cath IPPM Agenda IETF 102 When: 13:30 - 15:00 EDT, Wednesday 18 July 2018 Where: Saint-Paul/Saint-Catherine, Fairmont The Queen Elizabeth, Montreal Chairs: Brian Trammell, Bill Cerveny, Tommy Pauly Welcome, Note Well, Agenda, Status Chairs 2 documents in the RFC editor queue WGLC on twamp-test --------------------------------------------------------------- draft-ietf-ippm-metric-registry A. Morton draft-ietf-ippm-initial-registry A. Morton Al presenting https://datatracker.ietf.org/meeting/102/materials/slides-102-ippm-1-2-initial-performance-metric-registry-entries-draft-ietf-ippm-initial-registry-and-draft-ietf-ippm-metric-registry-00 Al - Summary (Slide 3 - 6) of metric and initial registry, no revisions to metric-registry Al - Brian Trammel has promised to provide comments on TCP metric Discussion points (Slide 7) Brian T> Random source port wont fit in, Use DNS ID (16 bits) - How about support for stress test with 16bits on DNS server Nalini > DNS is getting encrypted. Al > The measurements are at active so has access to DNS ID Al > UDP periodic one-way delay and loss, preference? [no opinions] Next steps (Slide 8) Rachel > XRBLOCK not sure of interest/how it will progress Tommy > Requests the wg for feedback --------------------------------------------------------------- draft-ietf-ippm-route A. Morton Al presenting https://datatracker.ietf.org/meeting/102/materials/slides-102-ippm-3-advanced-unidirectional-route-assessment-aura-draft-amf-ippm-route-00 Slide 3 > Routing class C does not refer to address class C. Slide 4 > Passive observation of a stream at transit/mid-point Brian T>Volunteers Jari for mid-point RTT/traceroute like tools to look at this document Brian T(chain hat on)> Intended status - standards track Al > Interested in determining what other measuremnts can be done at the mid-point vantage point Al > Looking for review for Section 4 and beyond Alia volunteered to review the draft --------------------------------------------------------------- draft-ietf-ippm-ioam-data F. Brockners draft-brockners-ippm-ioam-vxlan-gpe draft-brockners-ippm-ioam-geneve draft-weis-ippm-ioam-gre draft-ioametal-ippm-6man-ioam-ipv6-options draft-spiegel-ippm-ioam-rawexport Frank B presenting https://datatracker.ietf.org/meeting/102/materials/slides-102-ippm-4-to-8-in-situ-oam-ioam-numerous-ioam-drafts-00 Summary of changes 02 -> 03 : Layering, Security consideration Open issue > No namespace for the trace data collected to qualify the data such as app-data, node, interface IDs, Proposal > To include namespace ID and carry namespace specific flags/data, to resolve limitation of IOAM data field definitions. Encap drafts > reuse of formats from IOAM data draft - VXLAN-GPE, GRE, GENEVE, IPv6 GRE can be used to carry IOAM in IPv4. IPv6 encapsulation new draft - defines HbH and DO IPv6 options to carry IOAM data Mickey Speigel presenting Raw Export of IOAM Data using IPFIX Brian T (not as chair)> As a mainatiner of IPFIX namespace registry - The current proposal reusing the sections for IOAM raw export will break Collectors. Consider defining new elements for IOAM packet type. Create new ones review https://tools.ietf.org/html/rfc7013 The ipheaderpacketsection, datalinkframesection were defined and optimized for PSAMP Nalini > 1. Functionally this is cool. 2. Enterprise deployments capture entire packets. why not do that? Mickey > Helps in filtering, optimizing data Brian T relaying yakob stein > Privacy/security wire tapping issues, similar to during PSAMP standardization Frank & Mickey > Next steps Brian T > Any objections to WGLC for data draft Briant T > Update the draft with the namespace and then ask for WGLC Alia Atlas > Packet format for GRE and Geneve need to review. On the namespace - likes the idea, lets you do additional administrative policies Mickey > Encap drafts - IPPM wg adoption? Mirjam > Encap drafts not in charter? Brian T > Idea is to start the encap & review in IPPM and then take it to respective wgs. Has it been discussed in encap WGs? Frank > Geneve (NVO3) discussed, NSH yes, GRE no, 6man no. Tommy > Create awareness about the encap in the respective WGs before adopting it here. Alia Atlas > Good to get it review in RTGWG. VXLAN GPE is interesting, no RFC for it yet, will be informational eventually. You need an ethernet protocol code point. Mirjam > Havent reviewed the encap drafts. Prefers the encap drafts to progress in the respective WG rather than in IPPM. NEed AD, chair discussion. For 6man the work can progress in IPPM per process in 6man, other encaps should be discussed with the respective wgs. --------------------------------------------------------------- draft-fioccola-ippm-multipoint-alt-mark G. Fioccola Fioccola presenting https://datatracker.ietf.org/meeting/102/materials/slides-102-ippm-11-multipoint-alternate-marking-method-for-passive-and-hybrid-performance-monitoring-draft-fioccola-ippm-multipoint-alt-mark-01 Document changes, Summary Next steps WG adoption? Brian T> wont ask for WG adoption now, will consider after the queue of active drafts is drained sometime b/n IETF 102-103. After another WGLC --------------------------------------------------------------- draft-fear-ippm-mpdm N. Elkins Nalini presenting https://datatracker.ietf.org/meeting/102/materials/slides-102-ippm-10-m-pdm-draft-fear-ippm-mpdm-00 Combines marking method and PDM - as PDM is end to end, need path data Advantages of marking - batching, aggregating averages e.g. delay M-PDM as Hbh options Open to ideas/opinions on Middle box identifier Brian T (no hat) > Scope? single administrative domain? Nalini > PDM is Internet wide, this could be single administrative domain, undecided Brian T > Middle box identifier needs to be bigger if it is across domaina, integrity protection Brian T > Why PSN at each middlebox hop? Nalini > PSN is useful to find drops Frank B > Timestamp formats? Nalini > Like to find the interest in this work before fully fleshing it out. Open to collaborate with the IOAM folks --------------------------------------------------------------- Ligntning Talks (as time permits) draft-alavarez-hamelin-tictoc-sic-01 (J. Ignacio Alvarez-Hamelin) Ignacio presenting https://datatracker.ietf.org/meeting/102/materials/slides-102-ippm-11-synchronizing-internet-clock-frequency-protocol-sic-draft-alavarez-hamelin-tictoc-sic-00 draft-zhou-ippm-ioam-yang-01 (T. Zhou) Tianran presenting https://datatracker.ietf.org/meeting/102/materials/slides-102-ippm-13-a-yang-data-model-for-in-situ-oam-draft-zhou-ippm-ioam-yang-01 Request for review and adoption - https://datatracker.ietf.org/doc/draft-zhou-ippm-ioam-yang/ Brian T> would you like to trying an experiment to progress the model in github as the focus of work, treating the generation of drafts/RFCs as secondary (this is somewhat described in https://tools.ietf.org/html/draft-claise-semver-02#section-3). Authors open to the process *********** Couldnt get to the following lightning talks ***************** draft-song-ippm-ioam-tunnel-mode-00 (T. Zhou) draft-gandhi-spring-udp-pm (R. Gandhi)