Created attachment 68733 [details] Log of systemd with no restart after failure When a simple service experiences a watchdog failure due to lock-up, stop is never called on the service. The service is transitioned to the "failed" state, but no further action is taken. If "Restart: on-failure" is set, multiple instances of service will be started because the previously failed instances are not stopped. I've attached a simple test application that just pets the watchdog a few times and then stops. The attached log shows the resulting behavior for both the default case and the "restart: on-failure" case. Tested with a user instance of systemd v44 on F17.
Created attachment 68734 [details] Log of systemd with restart after failure enabled
Created attachment 68735 [details] Simple watchdog test application
Created attachment 68736 [details] Watchdog test service definition
This should be fixed in git now. Closing.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.