Home Assistant outage when running appdaemon-predbat

[rholligan]

Describe the bug I’m experiencing some issues with Home Assistant when running appdaemon-predbat add-on (I don’t experience these issues when I stop the add-on). I’m running everything on a Home Assistant Green. Predbat starts up without issue, generates a plan and runs for a couple of hours. Then Home Assistant becomes completely unresponsive, UI is not accessible through port 8123 (or Observer port 4357). Sometimes the system seems to recover, and other times I just perform a hard reset (Predbat start on boot disabled) The issue seems to affect the whole system, e.g. there is data outage from other entities (heating system) when Home assistant is locked up (see sample chart).

I’ve been having a look through predbat.log, and once system comes back online, there is always a few lines similar to these: 2024-03-13 20:31:52.602363 WARNING AppDaemon: Excessive time spent in utility loop: 130529.0ms, 130395.00000000001ms in check_app_updates(), 133.99999999998545ms in other 2024-03-14 10:54:42.494363 WARNING AppDaemon: Excessive time spent in utility loop: 1730222.0ms, 40099.0ms in check_app_updates(), 1690123.0ms in other

I’ve also noticed a few instances of the inverter clock getting out of sync – could potentially be related? I do have 2024-03-13 18:56:59.905151 INFO pred_bat: Invertor time 2024-03-13 17:11:54+00:00 AppDaemon time 2024-03-13 18:56:59.895295+00:00 difference -105.1 minutes 2024-03-13 18:56:59.915643 INFO pred_bat: WARN: Invertor time is 2024-03-13 17:11:54+00:00 AppDaemon time 2024-03-13 18:56:59.895295+00:00 this is -105.1 minutes skewed, Predbat may not function correctly, please fix this by updating your inverter or fixing AppDaemon time zone 2024-03-13 18:56:59.990789 INFO pred_bat: WARN: Inverter control auto restart trigger: Clock skew >=5 minutes 2024-03-13 18:57:00.010091 INFO pred_bat: Calling restart shell command: rm -rf /homeassistant/GivTCP/*.pkl 2024-03-13 18:57:00.225886 INFO pred_bat: Calling restart service hassio/addon_restart with addon a6a2857d_givtcp 2024-03-13 18:57:15.291712 INFO pred_bat: ERROR: Exception raised Auto-restart triggered 2024-03-13 18:58:35.342165 WARNING AppDaemon: Excessive time spent in utility loop: 5619.0ms, 5544.0ms in check_app_updates(), 75.0ms in other 2024-03-13 19:31:50.754579 INFO AppDaemon: Previous message repeated 507 times

Expected behavior Predbat to run in the background without impacting functionality of Home Assistant

Predbat version appdaemon-predbat version: 1.0.8 Predbat version 7.16.5

Environment details

• Inverter and battery setup: Givenergy 5kW hybrid inverter Gen 3 / 9.5kWh battery Gen 3
• Standard HAOS installer or Docker: standard (Home Assistant Green)
• Anything else?

Screenshots

Log file predbat.log https://pastebin.com/ZQTWhSEk Appdaemon.yaml https://pastebin.com/4CxTtwaY Apps.yaml https://pastebin.com/NUQX3w8k Configuratino.yaml https://pastebin.com/BLRw0fHr

[springfall2008]

Hmm it seems like it locked up, not sure exactly why that might be. Can you try changing to 'threads: 0' to disable multithreading? A value of 1 isn't a good choice, either use 2 or more or leave it disabled.

[rholligan]

Will give that a try. Threads was set to "auto" before when this issue first occured, I tried changing to "1" but didn't improve things.

[rholligan]

Has been running successfully overnight with threads set to "0", issue appears to be resolved. Thanks for quick response!