I am experiencing a bit of a weird issue on Homeassistant 0.69: the graph cards in the frontend are looking as if the sensor hasn’t produced any data after a while. If I go into Grafana to show the values from my influxdb then there clearly is data coming in. Also automations work flawlessly. What could be the issue here?
are you getting errors from History or Recorder in your logs?
I have quite some errors in my log, but nothing that appears to be connected to History or Recorder…
What are the errors?
See below a part of the log. I have tried numerous times to figure out the source but have never been able to pinpoint it. Biggest issue nowadays is that I quite often get the error “OSError: [Errno 12] Cannot allocate memory#033[0m”.” After which the whole system goes down…
2018-05-26 07:26:12 ERROR (MainThread) [homeassistant.core] Timer got out of sync. Resetting 2018-05-26 07:30:25 ERROR (MainThread) [aiohttp.server] Error handling request Traceback (most recent call last): File "/home/pi/.homeassistant/deps/lib/python3.5/site-packages/aiohttp/web_protocol.py", line 231, in data_received messages, upgraded, tail = self._request_parser.feed_data(data) File "aiohttp\_http_parser.pyx", line 295, in aiohttp._http_parser.HttpParser.feed_data aiohttp.http_exceptions.BadStatusLine: invalid HTTP method 2018-05-26 07:34:24 ERROR (MainThread) [aiohttp.server] Error handling request Traceback (most recent call last): File "/home/pi/.homeassistant/deps/lib/python3.5/site-packages/aiohttp/web_protocol.py", line 231, in data_received messages, upgraded, tail = self._request_parser.feed_data(data) File "aiohttp\_http_parser.pyx", line 295, in aiohttp._http_parser.HttpParser.feed_data aiohttp.http_exceptions.BadStatusLine: invalid HTTP method
Memory issues are not good. That means the system is trying to use ram to perform an action and there isn’t any ram available. What hardware are you running on? You may have a fluke situtation where your hardware is starting to fail at the same time as upgrading to 0.69. It may look like upgrading to 0.69 caused the issue but your hardware is just failing. You could try to downgrade but if your hardware truely is having issues, you’ll still have them.
I am running on a Raspberry Pi3. I do have to admit that I am running quite a bunch of sensors:
25 web-based sensors (like weather and Google Driving)
20 433mhz sensors
24 433mhz actuators
3 powerplugs via WiFi
2 powerplugs via USB
8 power sensors from smartmeter via USB
4 temperature sensors/actuators
Another 15 various sensors
Around 25 automations
…
Could the Pi be too low on specs to cope with this? Only other things running on Pi are Influxdb, Grafana and Mosquito.
Put the following in your sensors config and see what it says:
- platform: systemmonitor
resources:
- type: memory_use_percent
- type: swap_use_percent
- type: processor_use
I’ve found my database tends to corrupt quite easily (also running RPi3) and the front end graphs don’t seem to update, so I just delete all the database files, reboot and it’s back to normal again.
I also replaced the SD card on the last install and haven’t had any issues with that since.
1 Like
I moved the recorder DB to MariaDB, overclocked the SD-card and now also set the swap size to 1024 (instead of 100). Let’s wait and see if this solves the issue…