[Teas] Discussion of service path computation and TE path computation

[yuchaode <[e-mail]>]

Hi TEAS Working Group & Chairs,

In the meeting of IETF 118, I raised some concerns after Sergio's presentation of TE path computation draft.
On the meeting, Oscar asked me whether my scenarios had any impacts on the TE path computation model. Because of lack of enough discussion, my answer was not so sure and didn't give a clear judgement.
Yesterday, I had a meeting with some authors of TE path computation draft, including Sergio, Dieter and Italo, and we fully discussed the scenarios I raised.
Finally we reached these conclusions:

1.       The scenarios I raised are from a parts of Operators who prefer a top down approach. They would like to do a path computation from service level and more intend-based. These Operators only want to care about the configuration on the service level and let the domain controller to finish the calculation & configuration of tunnels underlay.

2.       The current TE path computation is more like a bottom up approach. People who like this approach prefer to do the calculation & provisioning layer by layer. Take transport technology for example, they would like to calculate and provision WDM tunnel at first then to OTN tunnel.

3.       These two approaches come from different kinds of configuration requirement, so they are not conflicted.

4.       It is suggested to define a new RPC in client signal draft of CCAMP to provide a top down approach of path computation. Though it will probably reuse a lot of attributes of TE tunnel, it is not needed to refine the TE path computation model.

So we considered that there is no impact on the TE path computation draft and it is also ready for LC of TE path computation draft.

Thanks
Chaode