Juergen Schoenwaelder <[email protected]> writes: > On Wed, Feb 24, 2016 at 10:30:40PM +0100, Martin Bjorklund wrote:
> In your I-D (if I got this right), you only declare mount-points in > the schema and then an implementation can mount whatever it likes on a > mount-point. What is the use case for this? Why is it a feature to not > express in the schema at design time what can be expected behind a > mount point? B/c that schema may not be written yet, or known about by the mounting model authors, etc. Think of the case we have been driving this with (i.e., a virtual router running inside a host router), in this case we want all the modules at the mount point which the virtual router has at it's root. The schema that defines this mount point (the logical-network-element schema) wouldn't pre-specify all known models, and additionally we do not want to require model authors to augment the logical-network-element schema to add their schema to it. We just want it to work, and by not restricting what can be mounted there (other than to say it mirrors the root of the LNE), it does. Thanks, Chris.
signature.asc
Description: PGP signature
_______________________________________________ netmod mailing list [email protected] https://www.ietf.org/mailman/listinfo/netmod
