Syncthing resets to default profile every few days

[SyncOrSwim]

Hi,

Not sure why, but my syncthing is regularly resetting itself to default settings and I have no idea why. It only does this on my TrueNAS node via docker deployment.

I’ve checked permissions, and they appear to be fine. So what gives?

[SyncOrSwim]

I should also note that the syncthing container is deployed with UID/GID 1000:1000 which is the same owner as the NAS user, and the user of my other systems it syncs with for the sake of smooth syncing.

Still, it resets like this occasionally and I see no suspicious logs as to why.

[bt90]

Looks like your config is not mounted as a volume and is nuked as the container gets recreated.

[SyncOrSwim]

Hi,

That’s interesting to hear. Do you see any issues with my compose?

services:
  syncthing:
    image: syncthing/syncthing:latest
    container_name: Syncthing
    healthcheck:
      test: curl -f http://localhost:8384/ || exit 1
    restart: unless-stopped
    security_opt:
      - no-new-privileges:true
    environment:
      - PUID=${PUID}
      - PGID=${PGID}
      - TZ=${TZ}
    volumes:
      - /mnt/StoragePool/docker/syncthing/config:/config:rw
      - /mnt/StoragePool/docker/syncthing/sync:/sync:rw
    ports:
      - 8384:8384
      - 22000:22000/tcp
      - 22000:22000/udp
      - 21027:21027/udp
    networks:
      - syncthing_net

networks:
  syncthing_net:
    driver: bridge
    name: syncthing_net

[bt90]

See syncthing/README-Docker.md at main · syncthing/syncthing · GitHub

I’d recommend mounting to /var/syncthing and also using host network mode.

[calmh (Jakob Borg)]

Yeah that’s not where the container stores it’s stuff.

[SyncOrSwim]

Thanks for your replies, it helped me narrow down the issue and where I went wrong.

I recently switched from the image: ghcr.io/linuxserver/syncthing image to the image: syncthing/syncthing version, and I forgot that they both require vastly different approaches to the networking bit of the compose.

It appears to work normal again.