Agenda
PCE Working Group Meeting – Thursday, July 23,
2015; 1:00-3:00 PM
Audio
stream: http://www.ietf.org/audio/ietf93/ietf93-congresshalli-20150723-1300.mp3
1. Introduction
1.1. Administrivia, Agenda Bashing (chair, 5 min)
1.2. WG Status (chair, 20 min) [25/120]
I-D gmpls-pce-extensions
- Last call comments were received from the Shepherd. A new version will be issued.
I-D inter-layer
- Related to the previous I-D.
I-D inter-area-as applicability
- Need to decide if we want to close document, or open to include other newer functions.
- Authors will ask the list as to how to best proceed.
I-D h-pce extensions
- Authors need to polish manageability considerations and security section.
- Authors would also like to see if the I-D should remain as Experimental status or move to Standards Track status if there is working group support and enough implementations.
2. Work in progress
and discussed on mailing list
2.1. PCEP Extensions for WDM Networks (Young Lee, 5 min) [30/120]
2.2. PCEP Extensions for Segment Routing (Jon Hardwick, 10 min) [40/120]
draft-ietf-pce-segment-routing
- Comment in room on code point assignment, observation that other working groups ensure TBD’s in IANA section are used until early allocation is requested.
- Clarification comment for SID depth Suggested that people look at the relevant SR I-D presented in IDR working group.
- Request for more discussion on the mailing list from the PCE chair.
2.3. Stateful PCE in GMPLS-controlled Networks (Xian Zhang, 10 min) [50/120]
draft-ietf-pce-pcep-stateful-pce-gmpls
- Clarification by the presenter that the work relates to GMPLS, and they will add some text that discusses relationship to MPLS networks. This specific discussion will also be taken offline onto the PCE mailing list.
2.4. YANG Model for PCEP (Dhruv Dhody, 10 min) [60/120]
- Suggestion from mic that if the TE model is abstract then
there may be more general applicability.
- Concern that the LSPDB should not be described in the PCEP
model as it’s not a PCEP specific.
- Authors ask TEAS Chair (Lou Berger) and previous Routing Area Director where the work should be initiated if the groupings are not adequate. General TE abstractions and groupings should be progressed in the TEAS working group.
- Further comment that anything PCEP specific should be PCE
working group. There may also be an argument to move generic YANG work to
NETMOD working group.
3. I-Ds not discussed on mailing list
3.1. Binding Labels (Jon Hardwick, 5 min) [65/120]
draft-sivabalan-pce-binding-label-sid
- Vocal support at mic from an operator for the work and importance of this work as Segment Routing (SR) is already deployed, and if a PCE is used in the future it must be possible to report the SID/biding labels used in the network to the PCE.
- Request to clarify if the work is related to only SR was clarified by the presenter. It would be possible to deploy this is in an “incremental” network where one side of the tunnel is SR and the other might be RSVP-TE.
- Question from mic asking what protection mechanisms have been considered? Presenter suggested that it would be agnostic to the PCE and dependent on what mechanism (Bypass, FRR, LFA, etc.) was used for the SR sessions themselves.
- Discussion for protection scenarios continued but due to time constraints the chair asked for the discussion to be moved to the list.
3.2. PCEP Security Discovery (Michael Wang, 5 min) [70/120]
draft-wu-pce-discovery-pceps-support
- Request from the PCE chair to discuss the performance impact of the proposal. Specifically would it really be an issue to use TLS without a discovery capability?
- Comment that advertising MD5 capability is not required, RFC5440 states MD5 must be used and therefore this I-D should be compliant with RFC5440.
3.3. PCEP Extension for Association (Ina Minei, 10 min) [80/120]
draft-minei-pce-association-group
- Comment from vendor that work is important and will be implemented by them, plus a number of other I-Ds are dependent on this I-D. Finally, the proposal will also be implemented in Open Daylight.
- A comment from Lou Berger that was not picked up by the mic or heard at the front of the room. Second comment from Lou as to why the authors are not following the Extended Association Object from RSVP-TE. Further observation that the Type field is far longer in RSVP-TE than what is proposed for PCEP.
3.4. PCEP Extensions for SFC Traffic Steering (Qin Wu, 10 min) [90/120]
draft-wu-pce-traffic-steering-sfc
- No comments.
3.5. LSP Scheduling (Yan Zhuang, 10 min) [100/120]
draft-zhuang-pce-stateful-pce-lsp-scheduling
- Question asking what happens when an LSP timer expires? The PCE will tear down the LSP.
- Request for authors to provide more motivation, requirements and use cases.
- Concerns raised as to how the PCC “schedule” will be synchronized between PCEs other PCCs.
- Further comments were made but were not picked up by the mic and heard at the front of the room.
- Chair reminded queue at the mic that discussion would have to move to the list which is open “24x7”.
3.6. Temporal LSP (Huaimo Chen, 10 min) [110/120]
- Request by Chair to discuss work proposal and address overlap with the previous LSP scheduling proposal.
3.7. PCE as Central Controller (Quintin/Dhruv/Luyuan/Robin, 10 min) [120/120]
draft-zhao-pce-pcep-extension-for-pce-controller
draft-zhao-pce-central-controller-user-cases
- Request by TEAS chair that authors investigate ACTN proposal and see if there is overlap and room for merging work items.
End of Session.