Documentation Grafana Cloud Monitor infrastructure Kubernetes Monitoring Monitor jobs

Monitor jobs

Monitor cron jobs and other job types

You can monitor manual jobs and scheduled (cron) jobs. Use the main menu to find and select All jobs. Use the Cronjobs and Jobs lists to view jobs across all Clusters and Namespaces, based on the time range you choose in the time range selector. You can view:

• A color-coded status indicator for each job
• How jobs are distributed and where jobs are placed across the infrastructure
• For cron jobs:
  • Last succeeded, to verify jobs are completing successfully
  • Last scheduled compared to succeeded, to view any gaps that reveal failed or skipped executions
• For manual jobs, Pods/completions to track when the job was run

Identify job types

The Type column on the Jobs list categorizes Jobs by their owner reference. Scheduled Job means the Job was spawned by a CronJob. Other types you’ll encounter:

• Manual Job: created directly with kubectl create job or an applied manifest, with no owner reference. There’s no automatic retry scheduling. After the Job completes or exhausts its backoffLimit, it stays until you delete it manually or its TTL expires.
• Operator-managed Job: owned by a CRD controller rather than a CronJob. Common sources are Argo Workflows (tasks), Tekton (TaskRuns), Spark Operator (SparkApplication), Flink, and Kubeflow. The controlledBy field points to the operator’s custom resource, not a CronJob.
• Helm hook Job: created by Helm during install, upgrade, or rollback with annotations like helm.sh/hook: pre-install. Helm may delete the Job after the hook runs, so these are short-lived and easy to miss in fleet views.

The monitoring implication differs by type. Scheduled Jobs have a CronJob you can inspect for missed schedules. Manual and Helm hook Jobs don’t, so stale Failed Jobs with no owner are the main signal to alert on.

To further investigate a job, click the job name to open its detail page.

Give it a try using Grafana Play

With Grafana Play, you can explore and see how it works, learning from practical examples to accelerate your development. This feature can be seen on the Jobs list.

Try it

On the job detail page, the Overview tab contains:

• Status, start time, end time, Pod status phase, logs, and events
• CPU and memory usage, to identify any over or under provisioning as well spot any gradual increases that indicate memory leaks
• Container logs for debugging failed runs
• Events for identifying error messages or unexpected behavior
• Runs table to track success/failure patterns over time, and understand duration and completion

You can further explore each job’s CPU and Memory tabs for greater insight.