Since about one week, my HA instance is super slow and I can’t figure out what the reason might be. Maybe someone can support me to find the culprit?
My setup is a Raspberry Pi 4 with 4GB, passively cooled (~42°C) sipping <2 GB of RAM and with a CPU usage of <10%. I have disabled every single add-on already, problem stays. It’s not that only retrieving recorder data is slow, but even every HA menu. Since about two months, I’m running on a NVMe SSD via USB (with MariaDB), so I thought it might be that, installed a backup on a SD card, same problem. When using a clean installation (no backup) the problem goes away. This leads me to believe that it is an issue with my installation, not with my hardware.
I’m running out of ideas here. Help! Which logs could I check out?
An example for my HA core log is attached here:
After jumping through hoop after hoop to be able to open the callgrind file, I managed it. Not really sure what do decipher out of it, though… Any idea?
I set the logger to debug and enabled debug on HA core and components, which tanked the performance even further. After weeks of debugging, I gave up. I bit the bullet and did a reinstall. While installing my custom components, I reached the point where everything all of a sudden became sluggish again: after installing this gpio integration. I was able to remove it from my backup and do a clean roll-back. Thanks everyone for the support!
