Monitoring · Our own fleet
Know the moment a machine stops
A watchdog for a fleet of water-cooled mining machines: it checks each one every minute and sends a WhatsApp alert when something goes wrong.
- OwnerOur own mining fleet
- ChecksEvery 60 seconds
- StatusRunning
The challenge
Every hour a miner is down is lost income, and the machines sit far from their owner. Their addresses change on every restart, so simple monitoring tools lose track of them.
What we built
- Each machine checked every minute and found by its worker name, not its changing address
- Network rescans at start, every 15 minutes and when a machine stops answering
- WhatsApp alerts for offline machines, low hashrate, internet outages and pool failover
- Many machines down at once reported as one power or internet alert, not a flood
- A daily report with uptime and average hashrate per machine
- An optional safe auto-restart mode with strict limits, then a call for someone on site
Where it stands
Running on the farm computer. Alerts that happen while the internet is down are queued and sent when it returns. Known limit: this machine model does not report water temperature, so that rule is ready but silent.