This document explains the different topologies and configurations you can have with Redis Data Integration (RDI).
Running the CLI on a separate host
This should be the best practice. The CLI is used by the RDI administrator.
In production, it is recommended that you install the CLI on a different host to the cluster or the Debezium Server.
This scenario requires:
- Copying the
application.propertiesfile to the host where the Debezium Server is running.
- Running the docker command while changing the volume argument (-v) appropriately.
This guide uses Docker to illustrate the required command. If you are using Podman, simply replace
docker run -d --rm --name debezium -v <ABSOLUTE_PATH_TO_APPLICATION_PROPERTIES>:/debezium/conf debezium/server
See README for specific instructions when using Oracle as the source database.
Running Debezium Server remotely to the source database
While it is technically okay to run the Debezium Server on the same host as the source database, it is not always the best practice.
- The database host needs all its resources to process queries.
- The database owners are not allowed to install a database client for an application on the database host. However, this scenario is fully supported and only requires a small update to the
Running Debezium Server remotely to RDI
In many cases, you wouldn’t want to run the Debezium Server on the same host(s) as the Redis Cluster node(s). In this case ensure you edit the sink section of
RDI and target Redis databases on different clusters
This can be done when you want a separation of concerns or you do not wish to install/enable modules on the target cluster. In this scenario the
config.yaml file needs to be edited with the following changes:
connections: # Redis target DB connection details target: host: <REDIS_TARGET_DB_HOST> port: <REDIS_TARGET_DB_PORT> password: <REDIS_TARGET_DB_PASSWORD>