[Teas] teas: covering in-network path calculation Q.

[Toerless Eckert <[e-mail]>]

Dear TEAS WG.

Hope newcomer questions are endured here on the mailing list:

I do understand how the TE architectures support the explicit 
offpath calculated path model. What i do not know if/how it
also describes/ deals with distributed or in-network path
computation options.

Browsing through the list of documents without trying
to read through all of them, it looked easier to ask.

The two well known options i am thinking of are:

- CSPF in RSVP-TE. Not distributed, but in-network, relying
  on IGP collected information.
- IGP metric engineering.

If/how are these considered in TEAs documents ? 

I am asking, because i am really a lot more interested in
path diversity engineering for resilience in DetNet
Unicast/Multicast or non-DetNet Multicast, and these are
today more obscure and likely not covered, but probably not
too difficult to add if the above ones have established
good proof points.

The two, non-offpath PCEC network assisted approaches
coming to mind are:

- disjoint path sets from disjoint instances
  of path calculation, e.g.: metric engineered IGP topologies.
  (yes, very problematic, i know. Just mentioning).

- disjoint path sets derived from a single instance of an
  algorithm, such as MRT.

Any explanation of current state of the art in TEAS for this
highly welcome!

Thank  you
    Toerless