Since the start of my HA experience journey I have had some issues with some integrations. That is these integrations more or less work flawlessly but the in the logging there is an issue.
Logger: homeassistant.helpers.entity
Source: helpers/entity.py:628
First occurred: October 16, 2021, 8:36:45 AM (2418 occurrences)
Last logged: 6:05:44 PM
Update of light.led_keuken is taking over 10 seconds
Update of light.led_tuin is taking over 10 seconds
Update of cover.switchbot_curtain is taking over 10 seconds
Update of sensor.work is taking over 10 seconds
I have the impression that these logs are flooding the log file causing HA to freeze up after a few days.
My question is can I in any way stop the logging of for instance the lights. And if so how do I do that?
My log file right now is 40 megabytes… and will typically be over 100 by the time i install an update, I do not think that has anything to do with your HA freezing (mine has never frozen)
Many people with freezing are using a raspberry pi, and presumably have either a poorly regulated power source, or the cpu is getting too hot, what hardware do you have
Using Argon One Forty M.2 with 256gb SSD. RPi 4 8gb and an original power supply. The power suply is hooked up to a zwave socket which measures in and output and that shows no irregularities.
Lets hope that the amber project doesn’t have those issues as it is based on the RPi board.
What I would like to try is stop the logging. But not quite sure how to do that?
What is your logger set to?, by default it is set to warning, you can always set that integration to “Error”
You should also have setup the system monitor so you’ll have a quick glance of the system info and also be able setup various automations based on the sensors.
I guess it’s set to default cause there is nothing set in the config.yaml.
I’ll have a play around with this. Basically I’d like to set the logging off for integrations that work well but have occasional glitches. like curtains, lights and shelly switches
The system monitor is running. and all sensors in that look normal.
Nevertheless HA crashed again yesterday, It seems that the memory use increases just before it does. I still have no idea why and how to solve it. It also seems it happens every 5 days or so.
What was the memory use at the time of the crash and the prior few minutes?
I think you have other integration issues that may be causing the crash, also what do you mean by crash?, what happens exactly?, what are the ERROR’s in the home assistant log?, are you using the built-in database or a custom database engine?, I’m only asking because I am also on a RPi 4 and never had a crash since upgrading from a RPi 3 back in May.
Also you may gain more traction/assistance from more advanced community members by providing more details of your particular installation type and hardware as well as any details on the issues experienced.
So I’ve installed HA on a RPi 4 with 8gb ram. Then running it in an Argon One Forty case with a 256 SSD.
Have a aetec 5 Zwave dongle but mainly use an existing box from Iungo (my energy monitering platform, which also contains a zwave module) to control the 5 Zwave devices in my instance. The cool thing is that these 5 switch can either be controlled through the Iungo platform or HA as they are setup as binary switches in with template through mqtt in to HA. This makes for the added bonus that when not home and experiencing a freeze I can reset HA by switching the case off and on. (while I am home that switch serves to startup my profitec in the morning ;-))
I have a cc2531 dongle running the ZHA integration. Not my favourite but most of the Zigbee devices work most of the time. Occasional reset.
There are indeed some errors and warnings in the log file. The most common are LED lights that have a zwave switch and behind it a LED wifi controller. The zwave swicth funtions as a way to manually and by means of HA turn it on and off. The LED controller serves as a means to change color. While the zwave switch cut the power HA keeps pinging that LED controller which in turn increases the log file.
To answer the first question last. The memory in creases and then HA becomes unresponsive. Can’t reach and basically fails to react. When looking at the router even the IP address is not shown anymore.
What I might have to do at some point in time is start from scratch Which is really something I hate doing as I’d have to redo configure add some 50 or 60 integrations.
I installed originally on RPi 4 4GB with an 64gb Micro-SD card. Via downloading an Image and installing that to the Micro is SD. That process I repeated after upgrading my RPI to an RPi 8GB. Using Backups to to reinstall HA.
I followed that procedure again when upgrading the too a case with SSD. Which all went fine. The upgrade had some issues with ZHA because apparently there is some interference with ZHA and SSD disks. Solved by an extension cable. Another thing with this case was, that bluetooth was not working in the Argon One case until I installed the I2C option for the case. Also solved.
The warning and errors however have been present since the beginning and seem unrelated to hardware being used for HA instance.
As mentioned before the is a steady increase of memory usage as well:
