Relational databases are used for several aspects in Vidispine Enterprise MAM:
Stores all data related to authentication and OpenIDConnect (OIDC):
Caches the information of the ConfigPortal GIT repository for faster retrieval by ConfigPortal UI and backend services.
Storage all information managed by VidiCore .
Provides the storage layer for the ActiveMQ message bus.
VidiFlow Camunda Broker (new in VidiFlow 21.1)
Holds data for all currently executed service and rule tasks that the VidiFlow Camunda broker has received from the Camunda engine. It replaces the in-memory data the broker used in VidiFlow up to 20.2, so broker can now be redundant and won't loose data on restart.
The data size per row is quite limited but scales with the number of tasks being currently executed. Task data will be removed from the table after broker has sent any completion message for the task to the Camunda engine.
The load correlates with the load of Camunda in the sense that every service and rule task will be created and deleted once. Additionally, it might be updated several times for lock extension depending on the task's execution time.
VidiFlow Process Engine
Database for Camunda BPMN engine.
VidiFlow Background Jobs
VidiFlow File Watcher
VidiFlow Contract Usage
Database for the contract usage.
This follow the microservice approach where each service has its own, dedicated data storage. This helps keeping the services clearly separated and allows more flexible updates or changes of the internal data model of each service.
Supported database types:
Microsoft SQL Server
Postgres (requires version 21.1 of all Vidispine Enterprise MAM products)
For Microsoft SQL Server a dedicated SQL Server instance should be available.