How do I keep docker from filling up root?

[aaron-baff-ad-net]

So I have had a VM (with HA Proxy frontend) for a few years now, occasionally update the docker image to the latest. But docker keeps filling up my root dir (I have a separate partition I have assigned for Frigate data in the config) and I keep having to shutdown docker, prune everything, and rebuild/relaunch. It's crazy, and I need it to stop because I don't keep track of the system used, and I notice when I go to check Frigate and it doesn't work.

/media/frigate is the local host path to where to store stuff

version: "3.9"
services:
  frigate:
    container_name: frigate
    privileged: true # this may not be necessary for all setups
    restart: unless-stopped
    image: ghcr.io/blakeblackshear/frigate:stable
    shm_size: "256mb" # update for your cameras based on calculation above
    devices:
      - /dev/apex_0:/dev/apex_0 # passes a PCIe Coral, follow driver instructions here https://coral.ai/docs/m2/get-started/#2a-on-linux
      - /dev/apex_1:/dev/apex_1
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /home/frigate/frigate/conf:/config
      - /media/frigate:/media/frigate
      - type: tmpfs # Optional: 1GB of memory, reduces SSD/SD Card wear
        target: /media/frigate/cache
        tmpfs:
          size: 1000000000
    ports:
      - "5000:5000"
      - "1935:1935" # RTMP feeds
    environment:
      FRIGATE_RTSP_PASSWORD: "<PASS>"
      PLUS_API_KEY: "<API KEY>"

[inkeep-ai-assistant[bot]]

The issue is your tmpfs cache configuration is pointing to the wrong target. According to the documentation, the tmpfs cache should target /tmp/cache, not /media/frigate/cache(1).

Your current configuration has:

- type: tmpfs
  target: /media/frigate/cache
  tmpfs:
    size: 1000000000

This should be:

- type: tmpfs
  target: /tmp/cache
  tmpfs:
    size: 1000000000

(1)

The /tmp/cache location is where Frigate writes initial recording segments before they're checked, converted to mp4, and moved to the recordings folder(1). By pointing the tmpfs to /media/frigate/cache instead, you're likely causing cache files to accumulate on your root partition.

Additionally, verify your Frigate config file has the correct paths. The database should be at /config/frigate.db by default, but if you need to customize it, you can set it explicitly(2):

database:
  path: /path/to/frigate.db

(2)

After correcting the tmpfs target to /tmp/cache, restart your container. This should prevent Docker from filling up your root partition since the cache will be in memory and recordings will go to your designated /media/frigate mount.

📚 Sources:

1. Installation
2. Advanced Options

---

Have another question? Just tag @inkeep.

[hawkeye217]

The AI's assessment is correct here. You should use /tmp/cache since the type of the volume is tmpfs, a temporary in-memory filesystem.