Heartbeat monitoring for cron jobs and workers

Failures a health check never sees

Every one of these ends the same way: no ping. That is all a heartbeat needs to notice.

• Jobs that never start: the host was replaced, the crontab is gone, or the scheduler skipped the run
• Jobs that hang on a lock or a network call with no timeout
• Jobs that fail because a secret rotated or an environment variable is missing
• Workers and queue consumers that crash and never restart
• Errors that go nowhere because output is piped to /dev/null

One call at the end of the job

Zero-dependency SDKs for TypeScript and Python sign each ping for you. Anything that can make an HTTPS request works too.

import { Zvono } from "@zvono/sdk";

const zvono = new Zvono({ secret: process.env.ZVONO_SECRET! });

try {
  const processed = await syncOrders();
  await zvono.ping("hourly-order-sync", {
    status: "up",
    metadata: { processed },
  });
} catch (err) {
  await zvono.ping("hourly-order-sync", {
    status: "down",
    reason: "sync-failed",
  });
  throw err;
}

import os
from zvono import Zvono

zvono = Zvono(secret=os.environ["ZVONO_SECRET"])

try:
    rows = export_report()
    zvono.ping("hourly-report", status="up", metadata={"rows": rows})
except Exception:
    zvono.ping("hourly-report", status="down", reason="export-failed")
    raise

Setting up crontab, Kubernetes CronJobs, GitHub Actions or Vercel cron? See how to monitor cron jobs, or the ping API reference.

Frequently asked questions