Usage#

Below are some example commands. Additional details are available using the --help option.

# Run the daemon (typically in the background)
ppm daemon

# List running services
ppm info
ppm ls

# Add a new long-running service
ppm add --name my_test_service --env 'RUST_LOG=trace' -- \
  sh -c "while true; do echo world; sleep 30; done"

# Add a one-shot service
ppm add --name my_oneshot_service -- ls -la

# Remove a service
ppm rm my_oneshot_service

# Get statistics about running services
ppm stats

# Generate a configuration file
ppm show-configuration > ~/.partner-pm.yml

How Does It Work?#

PPM spawns and monitors services, each of which can be in one of the following states:

• Created Temporary state; the daemon has not yet handled the newly created service.

• Running The service is live and executing.

• Finished The service has terminated normally with exit code 0, or received a SIGTERM.

• Stopped The service is paused after receiving a SIGSTP.

• Crashed The service terminated with a non-zero exit code, or received a signal other than SIGTERM.

When a service enters the Crashed state, it is automatically restarted by the daemon using an exponential backoff strategy: interval * (2^(nb_restart - 1))