Updating and upgrading highly available systems

The Pega Platform supports the separation of rule tables and data tables into separate database schema for the purpose of minimizing the downtime of production during updates of applications, frameworks, and the Pega Platform itself. Use split schema when performing updates that range from staging environments to production environments, as well as to facilitate rolling restarts on production servers.

The schema composition is logically composed of both:

  • A rules schema, containing:
    • The rule base
    • System data
  • A data schema, containing:
    • Data objects
    • Work objects

If split schema is deployed in production, Pega Platform updates with zero down time are possible. Organizations seeking to use rolling restarts and split schema to perform minimal or no down time upgrades of internal frameworks and applications must plan for production data management.

To ensure that data is migrated to the production system:

  • If new rules and data are backward compatible, import the application or framework RAP into production after performing the rolling restart. Any new content in existing data tables and changes to work queues are visible immediately.
    • To make new rules available upon next login, configure Authorization/RSLUpdateBehavior = fixedreq.
  • If new rules and data are not backward compatible, import the application or framework RAP into production after performing the rolling restart at a period of low to no activity. New rules, work queues, and data tables are visible immediately upon next login.