For self-monitored systems that "chug along" in the background it's also nice to have a Dead Man's Switch style monitoring: set up the bash watchdog to send reg

For self-monitored systems that "chug along" in the background it's also nice to have a Dead Man's Switch style monitoring: set up the bash watchdog to send regular "I'm still alive" messages to an external service. The external service then notifies you whenever these heartbeat messages stop (bash script crashes, system runs out of disk or RAM, system loses network connection, VM gets accidentally terminated, etc. etc.). There's a number of monitoring services implementing this idea, and I'm also building one of them as an open source project – link in profile.