This is driving me crazy. If anyone can help, I’d really appreciate it.
This automation (code below) got triggered twice in the last week without anyone actually pressing the input boolean that’s supposed to trigger it. The worst thing is, the automation activates a really loud siren sound over the outdoor speakers → unhappy neighbors. It’s purpose is to be there in case someone is trying to break in the home; at that point, anyone with Home Assistant (just 2 of us) can press the boolean “on” and trigger it. But, this week I was at work and my wife/partner was at the gym, neither of us using HA, when it triggered… and it happened more than once about 1-2 hours apart. Neither of us had HA open or were using it. The only way to trigger this automation is to turn the boolean to “on”.
The only other relevant thing from the day was that our internet service provider was having outages; the internet was intermittently on / off / on.
@pedolsky thank u for your thoughts. Attached is a screenshot of the trace from one of the 2 instances when it fired - they appear identical except for the times. The changed variables is below that. As far as I can tell, it was just the input boolean that changed to on. That boolean is only on one card in HA. Nothing else has access to it.
As far as the uptime sensor goes, I’m curious to learn more. Is this what you’re talking about? Do I just add the uptime integration and then add the condition you suggested to my automation? And then, how does that help?
That’s a great idea! Thank u! One other thought/question… are you aware of any way (integration or other) to drill down on what user or device triggered an automation?
I thought I was losing my mind when some of my devices would take actions for what seemed like no reason. Turns out, some of the integrations will get confused and trigger the wrong device if the IP address of your devices end up changing when your router is rebooted. I changed every single device to a static ip address on my network and on the router/dhcp server, reserved that address and asigned it to only that one devce (I think what I said was rdundant - but on some devices you can set it on the device - so I do it where one would expect the ip address to be assigned by the dhcp server, but I also assign it on those devices as well) - and the problem never came back. Voila! Maybe not your problem but that is what solved it for me. I thought I was losing my mind. I ws very surprised that some of the integrations must be relying upon the IP address of a device istead of on it’s mac address which is what would really be the right wayt to code an integration…
This would be a really good thought, and I’ll heed your advice to reserve static IPs anyway. But this automation is composed of just one helper and one automation (no IP addresses). The actual devices (speakers in this case) are in the automation’s action, but there is also a push notification in the action, and it got triggered too. So, it looks like the whole automation got triggered (not just the devices).
Does the trace show anything - it wouid show what had triggered the change… Also in your startup for HA do you update or reset the input booleans or ther settings? Maybe HA got restarted? Also check under settings I think it is - “related” - anything else use that same boolean? Is it possiblke that switch loses and then regains access top your network and something happens then when it disconnects and/or reconnects?
The trace shows the automation was triggered by the input boolean titled “siren”. That boolean only takes part in that one automation. I’m not sure how to check whether my input booleans reset at restart. I don’t think they do because I’ve done many system restarts in the past and the automation has never been activated.