Server quickstart

The local server is the smallest managed deployment. It provides an authenticated web control plane, durable metadata, jobs, schedules, and run diagnostics while managing its PostgreSQL process locally.

Install and start local mode

The server needs Java 17 or newer. Extract the server archive and keep its home separate from any CLI installation:

curl -fL -o ReplicaDB-server-1.0.0.tar.gz "https://github.com/osalvador/ReplicaDB/releases/download/v1.0.0/ReplicaDB-server-1.0.0.tar.gz"
tar -xzf ReplicaDB-server-1.0.0.tar.gz
cd ReplicaDB-server-1.0.0
export REPLICADB_SERVER_HOME="${HOME}/.replicadb"
./bin/replicadb-server start local
./bin/replicadb-server status

On first start, the launcher creates the local state and asks for the initial administrator through a hidden prompt when bootstrap values are not supplied. Keep the server home and its keyring together when backing up the installation.

Sign in and create the first job

Open the API URL shown by the launcher, sign in with the administrator created during startup, and create a datasource before creating a job. A job refers to datasources and replication settings; it does not copy connection secrets into the browser response.

Start with a manual run, inspect its diagnostics, and only then add a schedule. Stop the server cleanly after a local test:

./bin/replicadb-server status
./bin/replicadb-server stop

Keep the boundary clear

Local mode is a durable single-node deployment. When execution must scale beyond one host, run api and worker as separate processes against shared external PostgreSQL and a shared keyring; do not combine them as start api local. The API serves the control plane, while workers claim and execute runs. The worker management endpoint remains private and is not a public UI.

Continue with server installation, distributed deployment, server configuration, server troubleshooting, and security and TLS.