Release v1.18.0

Stage Manager now knows the storage engine and the named sinks of a Production, and a Performance checks that its Production configures every sink its application uses.

Added

• Production takes database.kind: MongoDB (the default), PostgreSql or SqlServer, the SQL engines for an External database only (part of #99)
• Production takes sinks: default readModels and graphs names, and named configurations of kind MongoDB, PostgreSql, SqlServer or Neo4j, each with a connection Secret or, for Neo4j, the Production's graph (part of #99)
• Production sinks are validated with problems that name the sink: duplicate names, unknown kinds, a default naming a missing sink, Neo4j for read models and SQL or MongoDB for graphs (part of #99)
• Performance takes application.sinks, the named sinks a traditional application uses; with productionRef the SinksMapped condition reports whether the Production configures each one, and a missing sink blocks the Performance before anything is applied (part of #99)
• Stage Manager documentation describes Production storage kinds, named sinks and the Chronicle kernel configuration they map to (part of #99)