relation-data-design

Design Juju relation data bags by defining separate application and unit data scopes, enforcing leader-writes for app data, and following best practices for key naming and string data storage to ensure predictable multi-unit communication.

Application data in Juju relations is written exclusively by the leader unit for app-wide configuration, while unit data is written by each unit individually to exchange per-unit status or credentials safely across relation boundaries.

Handle secrets in Juju relation data bags by using guarded access patterns for sensitive information and leveraging leader-writes for application data, ensuring credentials and status are exchanged safely across relations without exposure.

Best practices for Juju relation data include using clear key naming conventions, storing only string data types, respecting size limits, and correctly handling the relation lifecycle to ensure reliable data exchange between charms.

Yes, this relation data design approach is applicable to cross-model Juju deployments, providing safe and predictable data exchange where per-unit and application-wide data must be managed across separate models.

Use peer relations for inter-unit communication within the same Juju application, and choose provider or requirer sides for cross-application data exchange, ensuring correct scopes and access controls for reliable integration.