Cron Jobs

Schedule recurring background tasks by defining HTTP-based cron jobs. Tasks run automatically after production deployments; no separate job runner needed.

Globe supports defining cron jobs directly in your globe.yaml file. These scheduled tasks run periodically against specified HTTP paths in your production deployment, helping you maintain consistency, reduce manual overhead, and scale reliably.

Cron jobs in Globe.dev:

• Execute HTTP requests to specific endpoints in your application
• Only run on production deployments (not preview or development)
• Are synchronized with each successful production deployment
• Run in UTC timezone only

How to Manage Cron Jobs

Once defined, cron jobs are visible in your Globe dashboard after a production deployment. If a job no longer appears in your globe.yaml, it becomes inactive and stops running until re-added.

1. From your Globe dashboard, navigate to your project
2. Navigate to Cron Jobs
3. View your defined Cron Jobs by Active and Inactive jobs

How to Define a Cron Job

Here’s how you can define a simple cron job in globe.yaml:

crons:
  - id: refresh_auth_token
    schedule: "*/45 * * * *"
    path: "/auth/refresh" # This is the application endpoint that Globe will call

This example schedules a POST request to your application's /auth/refresh endpoint every 45 minutes. The path value is the HTTP route in your application that Globe will call. For more details on cron syntax, see the unix-cron documentation.

Globe ensures that these cron-triggered requests are private by default, meaning only internal events can trigger them. To make an endpoint public (not recommended for sensitive paths), disable the privacy setting:

crons:
  - id: refresh_auth_token
    schedule: "*/45 * * * *"
    path: "/auth/refresh" # Globe will call this endpoint in your application
    private: false # This endpoint is now publicly accessible

How to Handle Cron Job Requests

Each cron job triggers an HTTP POST request to the specified path. You need to write a route to handle this in your application.

Here’s an example using shelf_router:

import 'package:shelf_router/shelf_router.dart';

final app = Router();

app.post('/auth/refresh', (Request request) async {
  // Your handler for the endpoint defined in globe.yaml
  // Logic to refresh the auth token
  return Response.ok('ok');
});

If your handler returns a status code outside the 2xx range, it will be displayed as an error on the Globe dashboard.

Cron Job IDs

The id field represents the unique identifier of the cron job. It is used to manage job states and view logs in the dashboard.

• The list cannot contain duplicate IDs
• Max length: 50 characters
• Allowed characters: a-z, 0-9, and _

If a job is removed and later re-added with the same ID, its log history will persist.

Best Practices

• Keep sensitive endpoints private (default behavior)
• Name cron jobs clearly (e.g., refresh_auth_token, db_cleanup)
• Use tools like crontab.guru to help design cron expressions
• Monitor jobs via the dashboard to catch issues early
• Consider logging actions/failures in cron handlers for easier observability

Next

Use the Globe CLI to deploy, manage, and inspect projects directly from your terminal.