Does the Alert integration survive restarts?

Configuration
1

Does anyone know for sure if the Alert integration survives a restart?

Ex:

the alert criteria state is met and the integration begins alerting at the specified interval. But after it has transitioned to that state HA is restarted. Then after HA has restarted the state is already in the alert state.

Does the alert state act as an automation trigger (needs a transition to that state and then stay that way in order to alert and continue alerting at the time interval) or does it act as a condition (the alert integration checks the state at HA restart and continues alerting until the alert condition clears)?

I’ve never thought about this until now.

2

Both? In my experience what happens after restart with alerts is this:

  1. it does not remember anything from the prior restart. Like if you acknowledged it then it forgets that and sets itself from the state of it’s watched entity again. If you set dynamic intervals then it starts over as if it just became on again.
  2. you miss the actual trigger. It’s as if skip_first is always true this time.
  3. it does send out notifications at its repeat intervals as long as it’s watched entity remains in the selected state and it’s not acknowledged.

This is kind of why I like alerts because they do essentially survive a restart. Yes it’s a little annoying they start dynamic repeat intervals over and forget they were acknowledged (although the second part can be handled by automation) but still more useful then an automation that just ends at a restart.

1 Like
3

Thanks for the info.

I’m ok with all of what you described.

I just wouldn’t be ok if they needed to re-trigger again like an automation does.

I originally used an automation with some restart error checking. But then switched to an alert.

But then I got thinking about the above and wondered if i needed to go back to the automation.

It sounds like I should be good. :+1:

1 Like