Why is my storage getting filled?

1

Hi,

Recently my storage is getting out of hand, meaning that somehow it is over 95% and i don’t know why. I found out that the database file is getting larger every day, now it’s currently at 15 gb…
I delete it, restart home assistant and looking good… but only for a few days. Can u help me figure it out whats happenin’?
Ty

2

Obviously you are collecting a lot of data…

But if data is not important for you, you can limit it to f.e. 7 days, after that it will start purging:

recorder:
  purge_keep_days: 7
  auto_purge: true
  db_url: !secret my_sql_url

(I myself use a mysql database on another machine to prevent exactly the same problem)

or even disable it (using a ‘virtual’ database):

  db_url: 'sqlite:///:memory:'
3

I think that default purge interval is set to 10 days (if you didn’t change it). So, database increase should stop raising after 10 days.
Either reduce that number, increase storage, or you can select which entities to include/exclude:
https://www.home-assistant.io/integrations/recorder/

4

Thank you both for answers.
Strange thing that a few weeks ago this wasn’t a problem at all. My storage was at 30-40%, and not growing.
So this purge thing… I didn’t change this before, it was set to default. Already did this last week (after some research) but nothing seems to change since then.

Can i increase storage from 32gb to more? Is it possible or 32gb is just a default size?

5

Did you add any module/entity, especially anything that updates/changes very frequently (like every second or so…) recently?
Regarding memory: i think that it’s always possible. Depends on how you run your HA and what memory size you have currently. If you run HA inside VM then you just stop HA, assign more memory and restart HA.
If you run in RPi i guess that procedure would be: stop HA, power-off Pi, make an image of current card, transfer this image to a larger card, run Pi and expand storage. I wouldn’t know exact procedure or commands, though, since i run my HA in VM, not on Pi and have currently assigned 64GB for HA, and it’s plenty. However, i also run long-term history with grafana and InfluxDB.