Minutes ROLL Working Group Meeting - IETF-71 Philadelphia - March 2008 ====================================================================== Thanks to Greg Bernstein and Martin Vigoureux for having taken the minutes. Agenda ====== 1) Agenda/admin (Chairs - 5mn) [5] 2) Working Group progress (Chairs - 10mn) [15] 3) Inter-AS Requirements for the Path Computation Element Communication Protocol (PCECP) - draft-ietf-pce-interas-pcecp-reqs-04.txt (Kenji/Raymond/Nabil - 10mn) [25] 4) Update on Path Computation Element (PCE) Communication Protocol (PCEP) draft-ietf-pce-pcep-10.txt (JP - 10mn) [35] 5) A Backward Recursive PCE-based Computation (BRPC) procedure to compute shortest inter-domain Traffic Engineering Label Switched Paths draft-ietf-pce-brpc-07.txt (JP - 5mn) [40] 6) Extensions to the Path Computation Element Communication Protocol (PCEP) for Route Exclusions draft-ietf-pce-pcep-xro-03.txt (Eiji - 5mn) [45] 7) PCE applicability to WSON Routing and Wavelength Constraints draft-lee-pce-wson-routing-wavelength-01.txt (Young - 10mn) [55] 8) PCE GCO draft-ietf-pce-global-concurrent-optimization-01.txt (Young - 5mn) [60] 9) "Requirements for GMPLS applications of PCE" draft-otani-pce-gmpls-aps-req-00.txt (Tomo Otani - 10mn) [70] 10) Update on draft-ietf-pce-disc-mib-02.txt (Emile - 10mn) [80] 11) Discussion on rechartering (All - 10mn) [90] Minutes ======= 3) Inter-AS Requirements for the Path Computation Element Communication Protocol (PCECP) - draft-ietf-pce-interas-pcecp-reqs-04.txt (Kenji/Raymond/Nabil - 10mn) [25] Adrian> Inter-AS draft -- Adrian: Weak security section; Adrian wrote a block of text; New revision with new security section will last call end of this week. 4) Update on Path Computation Element (PCE) Communication Protocol (PCEP) draft-ietf-pce-pcep-10.txt (JP - 10mn) [35] JP> Last call just ended. Interoperability testing some public some private. About 10 implementations. Ready for publication request. This is a major milestone for the working group. 5) A Backward Recursive PCE-based Computation (BRPC) procedure to compute shortest inter-domain Traffic Engineering Label Switched Paths draft-ietf-pce-brpc-07.txt (JP - 5mn) [40] JP> ID has been stable for a while and implemented. IANA section has been updated in the latest revision: please review. Ready for Last call. 6) Extensions to the Path Computation Element Communication Protocol (PCEP) for Route Exclusions draft-ietf-pce-pcep-xro-03.txt (Eiji - 5mn) [45] Eiji> After the next revision (changes on AS sub-object) the draft will be ready for Last Call. 7) PCE applicability to WSON Routing and Wavelength Constraints draft-lee-pce-wson-routing-wavelength-01.txt (Young - 10mn) [55] JP> Question duration of light path are you calling for stateful? Young> not necessarily. This is more concerned with priority. Don O'cconnor> What do you mean by transparent optical network? Young> O/O/O rather than those with OEO. Don> I'm opposed to defining new networks here, that should be done at ITU. The optical networks you are refering to are not yet standardize by ITU. Adrian> A MEMs is an O/O/O switch so we are not covering any data plane here. Don> Restating same issue Young> This can include wavelength conversion. Igor B.> The model concerning one computing paths the other assigning wavelengths doesn't make sense to me. Greg> On RWA computation single or multiple path The calculation is about comparing paths because algorithm may be complex Might do k-shortest paths and then do graph colouring Middle model between distributed and multiple paths is looked for Impairment techniques may vary, but all have a minimal subset of continuity and lambda availability constraints Agree impairments is important Please see the literature common practice to compute multiple paths per source destination then apply approximate graph coloring algorithm. See reference in draft for an overview. Eve> Agree that points on computation are generic and... ? What impairments can be measured etc are difficult and are part of the domain of competance of other bodies Q6/15 and Q7/15 optical components and subsystems. Igor> RWA should be always done on same PCE otherwise risks of links shortage Emmanuel> Do you take into account the regeneration constraint? Young> Here quite generic, more specificities in CCAMP this work does not address physical constraints here. Malcolm> Have you considered taking into account work on impairment from Greg B? Young> For the moment not considering that. needs to mature. Malcolm> Should not neglet impairments from the biginning Young> We are aware, point taken. Eve> Physical layer experts are in Q6/Q7 in ITU Malcolm> What about power balancing and rebalancing Young> Not yet taken into account Malcolm> These are not constraints, these are control requirements Malcom> How are you going to communicate this work with ITU-T? Adrian> This working group is concerned with information between PCCs and PCEs. This working group is addressing protocol between PCE and PCC if people want to have support for externally defined functions (or externally to be defined functions) then we will ask them for a reference. So what to liaise, to communicate? Don> Agree with Adrian said no issue as long as you stick to what exists. 8) PCE GCO draft-ietf-pce-global-concurrent-optimization-01.txt (Young - 5mn) [60] Young Lee> Waiting for Chairs review. Chair's acknowledge. JP> Note on code points. Adrian> Request to editor of PCEP draft to send an up to date list of code points. 9) "Requirements for GMPLS applications of PCE" draft-otani-pce-gmpls-aps-req-00.txt (Tomo Otani - 10mn) [70] no question. JP> Who has red the draft? 2 people 10) Update on draft-ietf-pce-disc-mib-02.txt (Emile - 10mn) [80] requests WG LC for the MIB DISC Anybody implemented this MIB? No Answer. Adrian to Ross: How to progress this? Ross> I will look at it and come back to you on this. 11) Discussion on rechartering (All - 10mn) [90] JP> We first need to close action items pending. Who would like to initiate applicability statement? no answer. Items for rechartering: PM2P (not explicitly in charter today). JP> Who would be in favour? Vast majority JP> Who would be opposed? nobody JP> We will take it to the mailing list. Greg> WSON: should we consider something that we could standerdize? Adrian> Are you asking for a standard protocol for distributing TE information? Greg> yes Igor> We are very far from standardizing Adrian> Should look at multi-instance (ospf & is-is are doing that) Lou> Second Adrian comments multi-instance seems just the right thing for this Meeting closed.