Hi all

I am designing a framework for historisation implementation (SCD). One feature I would like to provide is a table in that the actual state of an entity is put and if this is complete, this history table is "updated":

      ------------         -------------
==>  | ENTITY_ACT |  ==>  | ENTITY_HIST |
      ------------         -------------

I plan to use instead-of-triggers on the hist table that read the actual table and perfoms all necessary inserts und updates on the history table. If I want the termination of a record version (actually the record of a specific business key with a specific payload) to get propagated up and/or down referential integrities (no overlapping validities) I have to make sure that only one of those processes is modifying a table. I was thinking of a scheduler queue where the trigger would put a process request and PostgreSQL would work through. Is there a scheduler within PostgreSQL? I read the documentation and searched the web but could not find a hint. But before going another road or implementing something myself, I ask. Maybe this design is no good at all.

Kind regards

Thiemo


Reply via email to