I have multiple 11.6G overlay files, and no listed commands flush them
Anybody has a solution?
ha su repair does not work and it is now taking 129gb…
1 Like
Experiencing the same issue, I’ve asked for support on Discord.
Did you get a reply on your question?
Same problem here!
Hi, I am having a disk space issue as well and I am suspecting it is this overlay2 folder taking up disk space. My DB is less than 400MB, backups folder is 360MB. df -h shows 11.3GB used, while du -sh shows just 1.3 GBs used.
[core-ssh /]$ df -h
Filesystem Size Used Available Use% Mounted on
overlay 23.9G 11.4G 11.3G 50% /
devtmpfs 978.6M 0 978.6M 0% /dev
tmpfs 980.6M 0 980.6M 0% /dev/shm
/dev/sda8 23.9G 11.4G 11.3G 50% /data
/dev/sda8 23.9G 11.4G 11.3G 50% /config
/dev/sda8 23.9G 11.4G 11.3G 50% /ssl
/dev/sda8 23.9G 11.4G 11.3G 50% /addons
/dev/sda8 23.9G 11.4G 11.3G 50% /backup
/dev/sda8 23.9G 11.4G 11.3G 50% /share
/dev/sda8 23.9G 11.4G 11.3G 50% /media
tmpfs 392.2M 1.0M 391.2M 0% /run/dbus
/dev/sda8 23.9G 11.4G 11.3G 50% /etc/asound.conf
/dev/sda8 23.9G 11.4G 11.3G 50% /run/audio
/dev/sda8 23.9G 11.4G 11.3G 50% /etc/resolv.conf
/dev/sda8 23.9G 11.4G 11.3G 50% /etc/hostname
/dev/sda8 23.9G 11.4G 11.3G 50% /etc/hosts
tmpfs 980.6M 0 980.6M 0% /dev/shm
/dev/sda8 23.9G 11.4G 11.3G 50% /etc/pulse/client.conf
tmpfs 980.6M 0 980.6M 0% /proc/asound
tmpfs 980.6M 0 980.6M 0% /proc/acpi
devtmpfs 978.6M 0 978.6M 0% /proc/kcore
devtmpfs 978.6M 0 978.6M 0% /proc/keys
devtmpfs 978.6M 0 978.6M 0% /proc/timer_list
tmpfs 980.6M 0 980.6M 0% /proc/scsi
tmpfs 980.6M 0 980.6M 0% /sys/firmware
[core-ssh /]$ du -sh
1.3G .
I run HassOS (supervised on a VM installed from the ova file). If I am not mistaken, that runs docker in the backend, the problem is I cannot access the /var/lib/docker folder, since it is not there. Also, I cannot run the docker system prune --all --volumes --force command. Can someone help me?
[core-ssh /]$ cd /var/lib/
[core-ssh lib]$ ls -la
total 28
drwxr-xr-x 1 root root 4096 May 31 20:51 .
drwxr-xr-x 1 root root 4096 May 31 20:51 ..
drwxr-xr-x 2 root root 4096 Mar 29 2023 apk
drwxr-xr-x 2 root root 4096 May 31 20:51 dbus
drwxr-xr-x 2 root root 4096 Mar 29 2023 misc
drwxr-xr-x 2 root root 4096 Mar 29 2023 udhcpd
[core-ssh lib]$ ha docker info
logging: journald
registries: {}
storage: overlay2
version: 24.0.6
Ran ha su repair but that did not free any disk space.
Also, I noticed my disk space started to grow after installing the ESPhome add-on, does this make sense or is it just coincidence?
Thanks!
You might be on so something concerning the relation between ESPHome install and the growing overlay2 folder. My disk size started decreasing after this install too!
From what I understand, some add-ons work by creating a new docker container to run from, and this is the case for ESPhome, so that new container is using ~2.5GB in my case.
The problem is we cannot see nor access those containers, since they are “hidden” by the HA supervisor, or that is my understanding. But I think there must be a way to purge them to reclaim some space back…
@frenck & @bdraco Hope we can we bring this persistent issue to your attention! If we can help by delivering logs or something, please let us know.
1 Like
following. Same problem here. Overlay2 consuming 15GB 
Same issue. Tried any suggestion from this and other threads but overlay2 keeps consuming the whole disk-space and even if I free some, the same problem reoccurs a few days later.
Is there anyone that can support us on this?
Same issue shows after updating supervisor to: 2024.03.0
At time graph, you can see exact moment when disk usage just jumped:
I have tried multiple options to fix that but with no luck so it has to do something with supervisor, can it just clean old data or something (yes, I have tried ha su repair but with no possitive results).
What I can see, each restart of docker or RPI generated new folders inside overlay2
UPDATE:
Interesting fact… I have removed all containers, cleared docker, overlay2 etc, installed everything from scratch (leaving homeassistant folder untouched) and overlay2 is again eating GBs of disk space…
ha su repair plus all the docker commands reduced mine from 34GB to 17GB. It’s good, but I would have loved more.
I experienced the same. Removing the esphome add-on and reinstalling it freed up around 6GB on my yellow
Have the same issue, overlay2 directory taking up 40GB.
docker system prune -a freed up a couple of GB only.
And ha su repair returned an error timeout exceeded.
I’m also having the same issues with disk space and I can’t solve them.
If I run ha su repair, after a while a timeout message appears, and at the next system reboot, this error shows up: “Error returned from Supervisor: System is not ready with state: setup”
Following this, I restored the previous system snapshot to restart it quickly.
Then I tried to run only the following commands, and once again, I encountered the same error: “Error returned from Supervisor: System is not ready with state setup”
docker image prune -a
docker volume prune
journalctl --rotate
journalctl --vacuum-time=1m