As was discussed on https://github.com/apache/trafficserver/pull/199, I've started the work to generalize this, and breaking it up in various parts.
This first Jira is about adding four things:
1) An internal process UUID, which goes into Machine. It's accessible from the core.
2) An API to get this Machine UUID.
3) A set of generic UUID APIs, to do things like stringification, Copy() etc.
4) A very simple API that exposes the HttpSM's sequences number (sm_id).