It is possible to add a Splitgraph engine as a replication client to a production PostgreSQL database, occasionally committing the changes as new Splitgraph images. This is done through PostgreSQL logical replication and has many uses:

• Recording the history of the upstream database for audit purposes
• Using anonymized production data snapshots for integration testing
• Building derivative datasets with Splitfiles as an alternative to data warehousing.

This example will:

• Spin up a PostgreSQL database with some sample data
• Set up a Splitgraph engine as a replication client
• Create an image from the data received from the origin database
• Make changes to the origin database, which will get propagated to the engine
• Record the changes as a new image

../run_example.py example.yaml and press ENTER when prompted to go through the steps.