Store and Forward example

This example displays a simple replication process using DistribuSys and Store and Forward, where a Store and Forward route between systems is established, and systems are configured to ensure redundancy between them.

In the figure below, Main System A is the owning system of the dataset, while B is the Store and Forward system. Back-up System B does not own any datasets, but it subscribes to the dataset owned by A. C and D are destination systems of the Store and Forward system that also subscribe to the dataset. The Test System E is configured with a direct route; normal replication occurs between A and E, as the dataset does not need to be forwarded from E.

Note: Before using the StoreAndForward table, doas_privilege records need to be set up for each route. These allow you to create subscriptions for data between the systems in the Store and Forward route.

In the example, a Store and Forward route between systems A & B is established, and Store and Forward routes between systems B & C and B & D are established. System B forwards data to systems C and D. It determines which system subscribes to the dataset and forwards it accordingly. In cases where system C or D has not subscribed to the dataset, the data is not forwarded.

Since there are no routes established between A and the two decision support site systems (DSS C and D), no data will flow directly between those destinations. If C or D makes a modification to a record owned by A, a command is forwarded from B to A. The number of direct routes are limited to reduce network traffic between the source and the destinations.

Note: If the doas_privilege table is modified, a new set of subscriptions will be generated for each system. The StoreAndForward table may need to be modified to reflect the new changes. As each route is activated, the subscriptions associated with the route are activated as well.