Call for help: Please help testing

[Tom-Bom-badil]

Heya,

looking for your assistance today - volunteers welcome. :)

Why?

Last week we got some reports that there is a serious bug (?) in UGOS: It stops updating some sensors whenever you are not connected through the UGOS Web Interface or the UGreen Mobile App to the NAS. As this is particularly related to CPU Temp, CPU Usage + RAM Load (probably amongst others), this is rather unconvenient for long-term monitoring.

What?

I have prepared a new container with an upcoming fix for this for version 2025.09.1 (instructions below, takes maybe 5 minutes).
The container is backwards compatible. So nothing should break (you never know).
I have tested this container on my 2 separate HA instances over the last 2 days now and in several scenarious - working well.

Advantages?

All your sensors should update properly (5s, 60s) after the new container is in place. Everything else remains the same.

Disadvantages?

Yes, one: If you run HA on a different machine than your NAS, you'll also need to restart HA every time after you have restarted the NAS. This applies only to the new funtionality - everything else should run as before, so you likely wouldn't even notice the change.

Risks?

It's software - there is always the risk that something is not working as before.
But I consider it to be low, as I tested this carefully, and UGOS runs more or less the same way on all NASes.
Also, you can always revert back to the main repo with a very few changes an clicks.

How?

• Please log on to you NAS, go to the Container and edit the Componse section as shown below.
• There will also be 2 new ENV variables in the future for log level (info, debug) and time zone (default 'Europe/Berlin' - needed to set the proper time in front of log entries).
• After editing, re-deploy the container and make sure you check to pull the latest version.
• Restart HA.
• Thats it.

Step1: Adjust settings

Step2: Pull and re-deploy

YAML Code for easy copy/paste:

services:
  token-api:
    # original --> image: ghcr.io/tom-bom-badil/ugreen_api_token_proxy:latest
    image: ghcr.io/tom-bom-badil/ugreen_api_token_proxy:develop-cca7bb0
    container_name: ugreen-api-token
    ports:
      - "4115:4115"
    environment:
      UGREEN_NAS_API_SCHEME: "https"
      UGREEN_NAS_API_PORT: "9443"
      UGREEN_NAS_API_VERIFY_SSL: "false"
      LOG_LEVEL: "info"       # default=info, only set to debug if needed.
      TZ: "Europe/Berlin"     # MUST be a valid timezone entry - relevant for logs only.
    extra_hosts:
      - "host.docker.internal:host-gateway"
    restart: unless-stopped

Thanks for helping! Please report back if you ran into any issues, but also make sure to let me know if you have any hints or remarks, or if everything went just smoothly without any issues.

/tom

[rexprojects]

Hey,

I think you need to remove the 3rd line with "build: ." - Had some issues with that and could not deploy.

Without it I was able to do it.

I also had to restart the integration once in HA but it seems to be working for now. Every few seconds there is an update so far.

Will monitor and keep you updated 👍

[Tom-Bom-badil]

Thanks @rexprojects, makes perfectly sense. :)

• The build: . was an relic from the old days. Removed accordingly, thanks for the find. :)
• Restarting HA: True - HA cannot know when the token_proxy is restarted. Added to the docs.
  In the past, the container was used very rarely (only on startup + whenever a token expired).
  This container is now working differently - it will keep a permanent connection to the NAS.
  But for that, it needs an initial 'ping' from HA with user name and password, because this is stored in HA, not in the container.
  v2025.09.1 will have a mechanism that pushes this information automatically HA --> Container every 60s, so whenever you restart the NAS, the container will have all necessary login information latest by 60s after startup.
  But for now, only a HA restart will/can trigger this.
  I'm planning to roll out the new version tomorrow, latest in 2 days (unless major bugs are showing up). Although the end user will notice hardly anything, there have been signifant changes in the 'engine room' over the past couple of days.

/tom

[rexprojects]

No issues detected since yesterday.

Sensors got regular updates, no issues in the logs.

Looks good :)

[Sineiko]

Von gestern abend 21Uhr, bis jetzt, scheint alles wunderbar zu klappen. Keine probleme bis jetzt.

[Danoro821973]

Läuft seit gestern Abend problemlos. Vielen Dank