Hi PCE, After some discussions with Dhruv about how and why we wrote RFC 8356, Haomian and I have posted a new draft to allow Experimental error codes in PCEP.
In summary, 8356 created space for Experimental PCEP messages, objects, TLVs. The assumption (see Appendix A) was that you could do anything experimental using these three elements, and nothing further was needed. This was true to some extent, but we have been concerned about the case of carrying Error-Types and Error-values in experiments. It could be possible for the experiment to define its own new object (say the "Experimental-PCEP-Error Object") and use that to carry experimental errors. However, we felt that that would make it harder to migrate an experiment to the standards track in the fullness of time. So this I-D creates space for Experimental Error-Types to be carried in the PCEP-Error Object. Note that it does not create space for Experimental Error-values within the existing Error-Types - we think that is unnecessary and would make the registry messy. The IETF datatracker status page for this Internet-Draft is: https://datatracker.ietf.org/doc/draft-farrel-pce-experimental-errors/ There is also an HTMLized version available at: https://datatracker.ietf.org/doc/html/draft-farrel-pce-experimental-errors-0 0 Please have a look and tell us if you think this is a good idea or a waste of time. FWIW, we believe we would use it in completing the work to move the PCEP-LS draft to be fully Experimental. Cheers, Adrian (and Haomian) _______________________________________________ Pce mailing list Pce@ietf.org https://www.ietf.org/mailman/listinfo/pce
