Agenda - PCE Working Group Meeting - IETF-68 - Prague - March 2007 Thanks to Martin Vigoureux for the Notes. 1) Agenda/admin (Chairs - 5mn) [5] 2) Working Group progress (Chairs - 10mn) [15] Good progress on several key documents (PCEP, IGP Discovery passed WG LC (still need to address some comments raised during the ISIS WG LC)). All protocol documents now have manageability section For information: inter-domain IDs passed LC in CCAMP Lot of work being done progressing fast and well while not much discussions on the Mailing list (ML): do not hesitate to use the ML. 3) Update on PCE MIBs (Emile Stefan - 5mn) [20] draft-ietf-pce-tc-mib-01.txt draft-ietf-pce-disc-mib-01.txt Module tags will be changed in the end. Change mostly for homing of three MIB modules. Notification impacts the MIB objects (need for threshold in table) Need to review that: not sure I propose good way. Potential DoS risk in case of notification from many agents Fix window instead of sliding JP: Would there be a sub-team of Service Providers reviewing the work and give feedback ? JLLR: ok JP: no need for a formal design team but it would be good to see more SPs providing feed-back on this work. Emile: not much work to do. Adrian: consider extensibility (as disco work progresses feature, capabilities we might have to extend MIB modules). Keep this in mind Emile: thinking an augmented field ? Adrian: yes for instance, would be good to plan this now. Emile: discovery or activity ? Adrian: actually everything 4) PCE communication protocol(PCEP) Management Information Base draft-kkoushik-pce-pcep-mib-00.txt (Emile Stefan - 5mn) [25] First version of PCEP MIB, not quite mature now Objective monitoring of PCEP Adrian: please do read these docs 5) Inter-AS Requirements for the Path Computation Element Communication Protocol (PCECP) - draft-ietf-pce-interas-pcecp-reqs-02.txt (Kenji - 10mn) [35] Slot cancelled. Next revision planned to be posted right after IETF-68. 6) Update on Path Computation Element (PCE) Communication Protocol (PCEP) draft-ietf-pce-pcep-07.txt (JP - 10mn) [45] JP: Several revisions since IETF-67. Good sign, starting to have implementations. New revision due to feedback of interop testing, clarifications, ... The base protocol is now stable Need to draw the line at some point and stop adding feature so that implementations can start and stabilize. No pending issues (objective function to be discussed later) Report of 5 existing implementations (FT, NTT, Cisco) Two other one, I do not know. Possibly share with community interop report in Chicago Dimitri: what type of implementations PCC, PCE, both ? JP: aware of a one interop with different implementation of the two sides. 7) Update on BRPC ID - draft-ietf-pce-brpc-04.txt (JP - 5mn) [50] Implementations of BRPC exist Now need a MIB. Volunteers welcome. Adrian: what goes into a brpc mib ? JP: cap. of downstream PC (to support BRPC) Adrian: restates extensibility of MIBs Dimitri: question on bidirectionality [...] JP: Please resend to the list. 8) A set of monitoring tools for Path Computation Element based Architecture draft-vasseur-pce-monitoring-02.txt (JP -5mn) Motivation: when doing testing, w/o. a tool do diagnostic, would be hard. All comments incorporated except most recent Dimitri : two important aspects * what are problems of combining implementations ? * what is exactly expected ? not overload thing before time JP: we're in line Dimitri : document still has too many metrics JP: We defined a flexible way to encode various performance metrics of interest. For example, if you take several sec to compute path you want to know where is the bottle neck. Adrian: seems draft needs some work and eventually pin point item to clarify material very pertinent to the WG. Would like one more rev before WG doc. JP: ok Dimitri, please bring forward items 9) Extensions to the Path Computation Element Communication Protocol (PCEP) for Route Exclusions draft-oki-pce-pcep-xro-00.txt (Eiji/Adrian - 5mn) [60] Next step: F bit in XRO No hand raised against becoming WG doc JP: ok seems we have good consensus, will take this to the list. 10) Encoding of Objective Functions in Path Computation Element (PCE) communication and discovery protocols draft-leroux-pce-of-00.txt (JL - 10mn) [70] Motivations: computation of one or a set of paths is subject to an objective function in RFC4674, it may be useful for a PCC to discover of supported by PCE other req in 4657 also Straightforward PCED and PCEP extensions Shall we defined objective functions ? 6 mandatory objective functions will be integrated in next revision. Adrian: how many have red ? hum quite a good number. Adrian : note that we do not do algorithms. JP: do we want to specify the objectives functions ? Proposal is to add them to this ID. 11) Path Computation Element Communication Protocol (PCECP) Requirements and Protocol Extensions In Support of Global Concurrent Optimization draft-lee-pce-global-concurrent-optimization-02.txt (Young Lee - 10mn) [80] Quite a few extensions to PCEP Igor: why make-before-break in the picture ? Young: for SLA sensitive type of TE, we want to avoid disruption. Igor: up to the signalling to decide JL: some may not need it, some must use it and it case of migration it is useful. Adrian: if you have set of TE LSPS before and after and if new ones use links of old ones you need to know which one you can tear down and those you can't Igor: ok thanks for the clarifications. JP: from an operational stand point a bit complicated. Do you plan an applicability statement and discuss pros and cons of GCO? I still have some concern about the scalability. Young: yes of course JP: aware of vendor planning to implement this? Young: yes. Adrian: you're quite right. We should look at it. Dimitri: what does global/local mean in charter? Is it linked single/inter domain ? JP: pretty clear : global means a whole set of TE LSP to reoptimize as opposed to a local optimal for a single TE LSP. 12) Carrying a Contract Identifier in the PCE communication Protocol (PCEP) draft-leroux-pce-contract-id-00.txt (JL - 10mn) [90] Straightforward PCEP extensions Comments to extend to a PCEP policy data object Will work on RSVP-TE extensions also to carry contract id Chairs : open to this initiative 13) Diff-Serv Aware Class Type Object for Path Computation Element Communication Protocol draft-sivabalan-pce-dste-00.txt (Jon Parker - 5mn) [95] Pce does not know class pool. Dimitri: in interprovider context, how do you assure global significance? Jon: this is an issue not tackled here. Adrian: how PCE has knowledge class pool ? You would suggest to build this knowledge based on IGP flooded information ? JP: please respin the draft tackling issue raised by Dimitri Not many people red the draft 14) Update on Policy IDs (Lou Berger - 10mn) [105] draft-ietf-pce-policy-enabled-path-comp-01.txt Adrian : Action point for chairs to do a thorough review groups please do also draft-bryskin-pce-pcpim-01.txt We think we completed the work.input & discussion is welcome Adrian: you think there is still more work to do? Lou: no we think we're done. Adrian: how many read the drafts (few hands). I would like a bit more interest. 15) Preserving Topology Confidentiality in Inter-Domain Path Computation and Signaling - draft-bradford-pce-path-key-02.txt (Adrian - 5mn) [110] AF : please object now if you do not want it a WG document. no objection raised 16) Requirements for Manageability Sections in PCE Working Group Drafts draft-ietf-pce-manageability-requirements-01.txt (Adrian - 10mn) [120] no post-presentation discussions