job-watcher is a Rust library for running and monitoring periodic
background tasks. It provides a simple framework for defining,
configuring, and observing jobs, and can expose a web-based status
page to visualize their health.
- Periodic Task Scheduling: Easily define tasks that run at regular intervals.
- Centralized Configuration: Configure task-specific properties like execution delay, out-of-date thresholds, and retry logic.
- Health Monitoring: Automatically tracks successes, failures, and whether a task has become "out of date" (i.e., not successfully completed within a configured timeframe).
- Web Status Page: Serves an HTML status page showing the state of all watched tasks, including recent output, execution counts, and environment details.
- Custom Alerting: Implement your own logic to determine when a task failure should trigger an alert.
- Stateful Tasks: Tasks can access their own execution history
(e.g., success/failure counts) via a
Heartbeatmechanism to make stateful decisions.
Here's how to get started with job-watcher
First, create a struct for your application and implement the
WatcherAppContext trait. This trait provides configuration and
top-level information to the watcher.
Next, create structs for your background jobs and implement the
WatchedTask trait. The run_single method contains the logic for
your job.
Here we define a simple, stateless task, and a second, stateful task
that uses the Heartbeat to fail after a certain number
Finally, in your main function, use the AppBuilder to register
your tasks and start the watcher. See the code under examples
directory for more details.