Monitor ETL pipelines and reporting jobs

The problem

Data jobs fail quietly and expensively. An ETL sync silently stops, a nightly report runs against half-loaded tables, and for days everyone trusts dashboards built on stale data. The cost isn't downtime — it's decisions made on numbers that were wrong.

How SteadyCron solves it

1. 1 Add a heartbeat check for each pipeline stage, matching its expected cadence.
2. 2 Ping /start when a job begins and /success when it finishes, so partial and hung runs are visible.
3. 3 Optionally include a short payload — row counts, a log tail — for context in the activity feed.
4. 4 Alert the data team on a missed or failed run, with consecutive-failure thresholds to cut noise.

TOKEN=<your-ping-token>
curl -fsS https://ping.steadycron.com/$TOKEN/start
rows=$(python run_etl.py) \
  && curl -fsS --data "loaded=$rows" https://ping.steadycron.com/$TOKEN \
  || curl -fsS https://ping.steadycron.com/$TOKEN/fail

Every job's status, schedule, and last run — at a glance.

Related documentation

• /docs/heartbeat-monitoring
• /docs/alerting