ETL job monitoring for missed data pipeline runs.

An ETL job that doesn't run usually doesn't crash loudly. The dashboard still loads. The CSV export still has data — just yesterday's data. The CRM still has customer records — just records that haven't been synced since Tuesday. Reports run on what's in the warehouse, and what's in the warehouse is whatever the last successful pipeline left there.

Data pipelines fail for predictable reasons that don't surface until someone notices the numbers feel off. A source API rotated its credentials. A schema change broke a transform halfway through. The destination database hit a connection limit and your job silently retried itself to death. A scheduled Airflow DAG was paused by accident and stayed paused for six weeks.

By the time someone says "these numbers look wrong", the gap is days or weeks deep — and the audit trail of when things went stale is buried in logs no one is reading.

Setup

Wrap any ETL run in one ping.

Whether the pipeline is a shell script, a Python job, or an Airflow task, the integration is one HTTP call after the work completes.

PingCron expects a check-in on the schedule you set (hourly, every 30 minutes, daily, etc.). If the pipeline stops checking in past the grace period, you get an alert.

Real-world examples

import requests

PING_URL = 'https://api.pingcron.io/ping/abc123'

def run_etl():
    extract_from_api()
    transform()
    load_into_warehouse()

try:
    run_etl()
    requests.get(PING_URL, timeout=10)
except Exception:
    requests.get(PING_URL + '/fail', timeout=10)
    raise

0 * * * * cd /opt/etl && python sync_warehouse.py && curl -fsS https://api.pingcron.io/ping/abc123

30 4 * * * /usr/local/bin/export_customers.sh > /exports/customers.csv && curl -fsS https://api.pingcron.io/ping/abc123

ETL and data sync jobs to monitor.

Any scheduled pipeline whose freshness matters.

Alerts where you'll actually see them.

Configure as many channels as you want per monitor.