streamProcessors need IDs to help identify them in checkpointing data. This ID should not just be the user supplied `name` of the streamProcessor. The ID can just be the ObjectID in the metadata DB, or something like `name-{random characters}`, or a UUID, etc. Choosing a unique ID allows us to distinguish when a streamProcessor of "name" has been dropped and re-created.

This ID and the tenantID needs to be supplied in the streams_startStreamProcessor request to mstreams.