#Clearing logs in Home Assistant Green (HAOS)?

1 messages · Page 1 of 1 (latest)

viscid chasm
#

I accidentally left a dev version of an app running in debug mode for a few days, and now my Green is very low on storage. I've deleted backups, I have no media, my recorder database is minuscule, and my Storage view shows that 23.7 out of the 24.3 GB used is 'system', so I'm 99% sure it's logs. But I don't have access to clear out the system logs via the Terminal & SSH app (as far as I know), and my other websearches have come to naught. Is there an obvious-to-everyone-else method of forcing a vacuum of the journalctl logs that I'm missing?

valid tusk
viscid chasm
#

Oh man, just having direct access is perfect. I was saying journalctl/vacuum because that's all I could think of that was left, but I was absolutely working blind. Thank you, now I can actually find the issue. 🙇‍♂️

#

Yup, that'll do it:

TYPE            TOTAL     ACTIVE    SIZE      RECLAIMABLE
Images          19        17        21.11GB   18.93GB (89%)
Containers      17        17        290.4MB   0B (0%)
Local Volumes   0         0         0B        0B
Build Cache     36        0         5.468GB   1.834GB```
valid tusk
#

Note that this does not show the bind mount data. HAOS doesn't usually use volumes. Better use the gdu way in my link for that.

#

Also ha supervisor repair does something similar to docker system prune.

viscid chasm
#

I ran the prune and now realize I was reading that docker system df output backwards, in that I'm now down to 17GB used with 100% of it reclaimable. I thought the "reclaimable" portion was the amount that could be pruned, but it was the inverse, so 11% in the screenshot above. But! I now have 8GB free and breathing room to figure out what I'm going to do going forward, in terms of giving myself more space. Thank you tons this has been incredibly helpful all around.

#

Oh, and gdu / showed my most-used space as /var at 517MB, which is basically all the journalctl logs, capped as you said.

valid tusk