Timeouts

Task timeouts allow you to limit the amount of time that a task will run before it is automatically cancelled. This can be useful to prevent "run-away" tasks from consuming excessive resources.

By default, all standard-runtime tasks time out after 1 hour, but this can be extended up to a maximum of 12 hours. The workflow runtime allows runs to execute for much longer—up to 60 days—but it does not yet support configuring a lower timeout.

In addition to timeouts, runs can be explicitly cancelled to stop it from completing:

When a run times out or is manually cancelled, it will be issued a SIGTERM and given 10 seconds to gracefully shut down. If the run has not yet exited after 10 seconds, then the run will be sent a SIGKILL.

Tasks can gracefully handle cancellation by listening for SIGTERM signals and performing any relevant cleanup before exiting on their own.

In the case where a run is cancelled before it is started, the run will be immediately marked as cancelled and the behavior above will not apply.

Timeouts are represented in seconds since a task was started.

typescript
Copied
1
export default airplane.task(
2
  {
3
    slug: "my_task",
4
    timeout: 600,
5
  },
6
  async () => {...}
7
);

javascript
Copied
1
export default airplane.task(
2
  {
3
    slug: "my_task",
4
    timeout: 600,
5
  },
6
  async () => {...}
7
);

python
Copied
1
@airplane.task(
2
    timeout=60,
3
)
4
def my_task():
5
    pass

yaml
Copied
1
timeout: 600