High availability

Watchdog (fencing)

Also called: hardware watchdog, /dev/watchdog, self-fencing, STONITH

A watchdog is a timer (often /dev/watchdog) that reboots a node unless it is regularly "pet" by a healthy process. On a Patroni primary, the agent pets the watchdog; if the agent hangs and can't demote PostgreSQL, the watchdog fires and reboots the node, forcibly removing the old primary before a new one can rise.

What this means

A hardware timer that will reboot a node unless a healthy process keeps petting it. On a Postgres primary the agent does the petting; if the agent freezes and can't safely step the old primary down, the watchdog fires and reboots the box, making sure the dead primary is truly gone before a new one takes over.

Why it matters operationally

This is the fencing mechanism that upgrades split-brain prevention from "probably safe" to "guaranteed." Without a watchdog, a wedged-but-alive old primary could in theory keep accepting writes after a new primary is elected.

← All glossary terms · GUC reference · Error catalog