Running a service

A service is a command that should keep running: a queue worker, a bot, a small server. RunWisp starts it with the daemon and starts it again when it exits.

runwisp.toml

[services.worker]
instances = 3                 # keep three copies running
run       = "node /app/worker.js"

Run runwisp reload and the three copies start.

Follow its output

runwisp logs -f worker

Each line starts with the copy it came from (worker#2 | ...). Every start of a copy is a run in the history, so when one crashes, its output and exit code stay behind for you to read.

Task or service?

If the command exits on its own when it’s done, it’s a task. If it runs until you stop it, it’s a service. A script that loops forever with a sleep inside is usually better as a task on a schedule; see A polling loop.

When it keeps crashing

A service that exits right after it starts is restarted with a growing delay. After too many failed starts in a row, RunWisp stops trying and alerts you like a failed task. See When a service can’t start.

Next steps

• Tasks and services compares the two in full.
• Every key a service accepts is in the [services.*] reference.