I run a business alone — my entire monitoring stack is one folder (3 files)

Description and use

A solo operator uses three simple scripts in one folder to monitor, log, and heal services without external SaaS tools.

Problem

Solo operators need a lightweight, reliable monitoring stack that doesn’t add maintenance overhead or cost.

Outcome

The stack detected a dead payment webhook in 5 minutes, a DNS expiry the day before it would have failed, and a scraper returning empty JSON for two weeks, all without external services.

Workflow

• 1. Run sensor.sh as a cron job to curl each critical URL and validate content.
• 2. Append each check result to ledger.txt with timestamp, target, and outcome.
• 3. If a check fails, healer.sh restarts the affected process, verifies the restart by checking port and endpoint, and logs the verdict to ledger.txt.
• 4. Review the last 20 ledger entries via a pager or web page for quick status.
• 5. Escalate or adjust as needed using the provided escalation ladder.

Evidence

A cron'd bash script that curls every money URL and validates the content, not the status code.

Source

One line per check: timestamp, what was checked, what was found. When something breaks at 3 AM you reconstruct the timeline in 60 seconds.

Source

Dead process? Restart it. Then prove the restart worked (check the port, check the endpoint). Log the verdict.

Source