Output Monitoring

Your cron job didn't fail.
It lied.

It ran. It exited 0. It pinged on time. And it exported nothing, backed up nothing, synced nothing. Most tools call that a success. DeadManCheck checks whether your job actually produced valid results.

Start free → See quickstart

Standard monitoring

✓ ETL job ran at 03:00

✓ Exit code 0

✓ Ping received

Rows exported: 0 · upstream API returned empty

Alert sent: no

DeadManCheck

✓ ETL job ran at 03:00

✓ Exit code 0

✓ Ping received

count: 0 · assertion: count > 0

⚠ Assertion failed — alert sent

The false success problem

Standard monitoring asks: did the job run? If yes, everything is fine. But that's the wrong question.

The right question is: did the job do what it was supposed to do?

⚠

Backup job ran but copied 0 files

Source directory was empty due to a misconfigured path. Exit 0. Ping sent. No backup taken.

⚠

ETL job completed but exported 0 rows

Upstream API returned an empty response due to an auth token expiry. The job handled it gracefully — and silently wrote nothing.

⚠

Sync job returned 200 but output was empty JSON

API returned {} instead of a populated payload. The script didn't error — it just synced nothing.

⚠

Nightly report ran but produced invalid output

A downstream schema change caused the report to generate malformed data — no errors thrown, no alert sent.

How DeadManCheck checks if your job actually worked

1. Send JSON output with your ping

# Bash

ROWS=$(run-etl-job.sh)

curl -s -X POST \

https://deadmancheck.io/ping/token \

-H "Content-Type: application/json" \

-d "{\"count\": $ROWS}"

# Python

result = run_etl()

requests.post(

"https://deadmancheck.io/ping/token",

json={"count": result.count}

)

2. Define assertions in the dashboard

Example assertions

count > 0

count >= 100

If any assertion fails

Alert sent immediately — even though the job exited 0 and pinged on time.

Available on the Pro plan.

Why exit code 0 isn't enough

Exit codes communicate whether a process ran successfully — not whether it produced the right output. Most scheduled jobs are designed to handle empty responses, missing files, or API errors gracefully — which means they'll exit 0 even when nothing useful happened.

Exit code tells you

• The process started
• It didn't crash
• No unhandled exception

Exit code doesn't tell you

• How many rows were processed
• Whether output was valid
• Whether anything was written
• Whether downstream systems got data

Checking results tells you

• Exactly what your job produced
• Whether it meets your expectations
• Whether to trust the result

Monitor cron jobs that succeed but do nothing

If you're searching for cron job succeeded but did nothing, job ran but returned empty data, or how to validate cron job output — DeadManCheck is built for exactly this.

The problem has a name now: false success. Your job exits 0, your monitoring dashboard stays green, and your data pipeline silently stops working. It's the failure mode that standard monitoring tools were never designed to catch — because they only ask "did it run?", not "did it work?".

DeadManCheck introduces a new check: send your job's output as JSON with the ping, set rules on what valid output looks like, and get alerted the moment your job produces nothing useful.

How output monitoring compares

A plain heartbeat monitor like Healthchecks.io only tells you whether a job ran and whether it ran on time — if your backup copied 0 files, it stays green. Cronitor can assert on a reported count too, but charges per monitor. DeadManCheck includes output assertions on unlimited monitors in Pro, and learns each job's normal duration automatically.

Capability	DeadManCheck	Healthchecks.io	Cronitor
Missing ping alert	✓	✓	✓
Duration monitoring	✓ automatic	Manual grace period	Manual only
Output assertions (count of work done)	✓ 1 free, unlimited on Pro	✗	✓
Unlimited monitors, flat price	✓ $29/mo	✗ tiered	✗ per-monitor
Open source	✓	✓	✗