Litestream

Command: start

v0.5.12 The start command resumes replication for a database that was previously stopped via the stop command. The database must already be registered with the daemon.

The start command communicates with a running litestream replicate process over the IPC control socket. The daemon must already be running.

Usage 

litestream start [arguments] DB_PATH

Arguments 

-socket PATH
    Path to the control socket.
    Defaults to /var/run/litestream.sock

-timeout SECONDS
    Maximum time to wait in seconds.
    Defaults to 30.

-json
    Output raw JSON instead of human-readable text.

Behavior

The start command is idempotent. If the database is already running (replicating), the command returns already_running status and exits successfully (exit code 0). This allows scripts to safely call start without checking the current replication state.

Output

By default, the command displays a human-readable confirmation:

status: started
db_path: /path/to/my.db
state: running
txid: 42
socket: /var/run/litestream.sock

With -json, the response is a JSON object:

Field Description
status Either started or already_running
db_path Absolute path to the database file
state Current state after command: running
txid Current transaction ID
socket Control socket path used

Examples

Start replication 

$ litestream start /path/to/my.db
status: started
db_path: /path/to/my.db
state: running
txid: 42
socket: /var/run/litestream.sock

Idempotent start

Calling start on an already-running database succeeds:

$ litestream start /path/to/my.db
status: already_running
db_path: /path/to/my.db
state: running
txid: 42
socket: /var/run/litestream.sock

JSON output 

$ litestream start -json /path/to/my.db
{
  "status": "started",
  "db_path": "/path/to/my.db",
  "state": "running",
  "txid": 42,
  "socket": "/var/run/litestream.sock"
}

Custom socket path 

$ litestream start -socket /tmp/litestream.sock /path/to/my.db

See Also

← Command: unregister
Command: stop →