Skip to main content

Cron Jobs

Cron Jobs are defined as console commands. The preferred way to run them is as a Kubernetes CronJob. Advantages of using Kubernetes Cron Jobs:

  • You can run multiple instances of the API in parallel without additional configuration for Cron Jobs.
  • Different resources can be used for Cron Jobs and the API.
  • You don't have to worry about blocking the API event loop.

Best Practices​

  • Test your Cron Job locally with the specified memory resource request.
NODE_OPTIONS='--max-old-space-size=256' npm run console demo-command

  • If you're using Kubernetes, avoid running a Cron Job every minute, as this will create a lot of overhead.

  • Don't run every Cron Job simultaneously (e.g. always on the first minute of an hour), as this might create a considerable load on all involved systems.

  • If you need alerting, consider using a Heartbeat monitor.

  • Ensure that your Cron Job allows repeated runs so manual runs are possible.

Cron Job Module​

A CronJobModule is available for @comet/cms-api which adds GraphQL queries and mutations for interacting with Cron Jobs. A CronJobsPage is available for @comet/cms-admin to display and manage Cron Jobs.

caution

If including the Cron Job module, ensure you have proper access control in place as this module allows interacting with the Kubernetes API.

Scoping​

The Cron Job Module respects the Kubernetes annotation comet-dxp.com/content-scope and will only allow access to the Cron Job if the user can access the specified content scope.

…
metadata:
    annotations:
        comet-dxp.com/content-scope: "{ \"domain\": \"main\", \"language\": \"en\" }"
…

Labeling​

Cron Jobs and Jobs can be given a human-readable label using the Kubernetes annotation comet-dxp.com/label.

…
metadata:
    annotations:
        comet-dxp.com/label: "Demo Cron Job"
…