Heartbeat Monitoring for Jobs and Workers

Heartbeat monitoring is essential for systems that cannot be validated with external HTTP checks.

Best targets for heartbeats

• Queue workers.
• Scheduled cron tasks.
• ETL and batch pipelines.
• Internal services behind private networks.

Implementation model

Each process sends a heartbeat on successful cycle completion.

Alert if:

• No heartbeat within expected window.
• Runtime exceeds historical baseline.
• Error count spikes between heartbeat intervals.

Practical thresholds

• Daily jobs: alert at 1.5x expected schedule interval.
• Minute-level workers: alert after 2 missed intervals.
• Long-running pipelines: combine start and completion signals.

Incident communication tie-in

When heartbeats fail, status updates should trigger automatically with clear impact summary.

Heartbeat coverage is a major part of a developer-first reliability stack:

Developer-first status pages playbook →