Node-RED unreliable the last week

Node-RED has started to be unreliable, it seems to affect call-service nodes, but not all. The most recent example is when I got home today. My arrival routine kicked off, and I was sent a notification saying so. This routine checks if anyone is home, and if not, turns on the thermostat, the kitchen lights, the Christmas tree, and opens the garage. Today, everything but the garage worked. Looking at the flow, the garage portion of the flow kicked off properly, according to Node-RED at least. The call-service node shows that the open cover service was called and then passed on to my telegram notification node, but the garage door did not open. Out of frustration, I restarted Hassio using the homeassistant.restart service (as this seems to correct the issue) before looking at the logs. Next time it happens, I’ll be sure to get thorough logs, but last time I looked I didn’t immediately see anything.

Until I get those logs, I wanted to see if anyone else is having any similar issues like this.

I did have the Node-RED integration installed from HACS, but have since uninstalled it to be sure that wasn’t causing the issue.

Here’s the info from my Node-RED addon -

-----------------------------------------------------------
 Hass.io Add-on: Node-RED
 Flow-based programming for the Internet of Things
-----------------------------------------------------------
 Add-on version: 5.0.7
 You are running the latest version of this add-on.
 System: HassOS 2.12  (amd64 / intel-nuc)
 Home Assistant version: 0.103.4
 Supervisor version: 193
-----------------------------------------------------------
 Please, share the above information when looking for help
 or support in, e.g., GitHub, forums or the Discord chat.
-----------------------------------------------------------

It’s acting very strange. The alarm was triggered this morning, not sure why, but the siren did not go off. Seems like pretty generic errors in the logs, just saying that that there was a call service error. If I restart home assistant though, they will start working correctly again.

25 Dec 07:10:23 - [error] [api-call-service:Siren on!] Call-service API error.  Error Message: 'triggered'
25 Dec 07:10:41 - [error] [api-call-service:Take Entryway Snapshot] Call-service API error.  Error Message: extra keys not allowed @ data['filename']
25 Dec 07:10:41 - [error] [api-call-service:Disarm House ] Call-service API error.  Error Message: 'triggered'
25 Dec 07:10:41 - [error] [api-call-service:Take Kitchen Snapshot] Call-service API error.  Error Message: extra keys not allowed @ data['filename']
25 Dec 07:10:41 - [error] [api-call-service:Take Garage Cam Snapshot] Call-service API error.  Error Message: extra keys not allowed @ data['filename']
25 Dec 07:10:41 - [error] [api-call-service:Take livingroom cam Snapshot] Call-service API error.  Error Message: extra keys not allowed @ data['filename']
25 Dec 07:10:41 - [error] [api-call-service:Take front door snapshot] Call-service API error.  Error Message: extra keys not allowed @ data['filename']

Here’s one of the call service nodes. AFAIK, this is properly configured.

Here’s one of the flows that gave the api error. It sends a snapshot of all interior cameras via Telegram when the alarm is triggered. I made no changes other than connecting the inject node and the flow worked.

I have no idea how to troubleshoot this weird issue…

I just noticed “Use alternative template tags for the Data field” is checked on the call-service node for my siren, not sure why. I’m not completely sure what this does as I’ve never used it, but I believe it’s if you’re using a template node to the call service node, which I’m not doing. I unchecked it, but obviously wont be able to test this until other people aren’t home, lol.

This was not checked in any of the other problematic call service nodes though.

Have u recently updated hassio? I had a bunch of API errors when I updated due to the naming of certain entity ids specifically my alarm nodes. I never put an entity id in previously but now it required one. After that it worked fine.

Do you mean like this? If so, I always use a static entity id, I’m not currently using any service nodes programmatically.

Yeah when I originally set p my alarm it never required an entity ID and worked fine because it was my only alarm, but since the update I was getting API errors and figured out that’s what was causing it.

There is definitely something going on, but I cannot figure it out.

I actually think it may be Home Assistant and not Node-Red at this point. I may take a snapshot of Node-Red and rest the rest of Home Assistant and see if that works.