Optimize Data Access Patterns To Update Job Execution Progress

A snapshot of metadata(job profile, etc) of a data import job, that is about to executed, is taken. This is stored in SRM and requested by other actors in the DI workflow(mod-inventory, SRS) to continue processing. This snapshot is taken from JSONB column in the database and can be quite large, at least 40KB. The database has to do string conversion of the data so this query takes longer than a regular query with relational columns. The request for this metadata is done multiple times over the course of a DI workflow so it is efficient to cache the metadata in the module instead of retrieving from the database over and over again.