Sorry for the scare, and thanks for reporting the issue with the screenshot, it helped me find the cause quickly.
Your subscription was never actually a problem: Stripe did have your trial running, and there was nothing to cancel or restart. The bug came from Gladys 5.0.4, which introduced a new mechanism: when Gladys Plus rejects a call due to « payment required, » the instance pauses the Plus features and displays the red banner. Except that the Gladys Plus server was returning exactly the same response in another case, unrelated to payment: when an instance requests a feature not included in its plan (e.g., backups, reserved for the Plus plan). Since Gladys checks the list of backups every night, your instance locked itself, and the « Check Subscription » button made the same call, so there was no way out.
The fix is on the Gladys Plus server side, there’s nothing to update on your end. Once deployed, the banner will disappear at the next automatic check (every night around 2 a.m.), or immediately by clicking on « Check Subscription » in Settings > Gladys Plus. Everything else (remote access, tablet, etc.) was never cut off: only the banner, and the Plus features already outside your plan, were affected.
And thank you very much for switching to the paid subscription despite this false start. It’s this support that allows Gladys to keep moving forward, it really matters.
Thanks for the feedback, and sorry: I was wrong in my previous message when I told you that the banner would disappear on its own. The server-side fix does prevent the issue from recurring, but it is not enough to unlock an already locked instance. Once the banner is displayed, Gladys only removes it if Gladys Plus confirms the subscription via the backup list, but your plan does not include backups, so this confirmation can never arrive. That’s why « Check Subscription » does nothing: the button makes exactly the same call.
The fix is now on Gladys’s side, in the next update: it removes this lock on startup, and the banner will disappear as soon as it is installed. Nothing to do on your end until then, and no need to click on « Check Subscription » again. As you noticed, everything else works normally.
Thanks again for insisting, it allowed me to see that the first fix was incomplete!
I’m still having the same display issue. I noticed that an update is available. I need to go from 5.0.4 to 5.1.1 but it’s not working
"Operations
Gladys v5.1.1 is available
Gladys will update automatically if you use Watchtower. If you want to update manually, click the button below. This will launch a Watchtower container that will only update the Gladys container; the other containers on this machine will not be affected.
No update has been applied: Watchtower found no new image for « gladysassistant/gladys:v4 ». Try again in a few minutes; the new image may not yet be available for your architecture. I restarted Gladys but that didn’t work either.
Thanks for the feedback! For the banner, it’s normal that it’s still there: the fix is in the 5.1.x version, and you’re still on 5.0.4. As soon as the update goes through, it will disappear.
For the update, I checked: the image gladysassistant/gladys:v4 is indeed up to date on Docker Hub (it has been pointing to 5.1.1 since September 22nd, for your architecture as well). The problem is therefore not coming from there, despite what the message says. I think rather that Watchtower failed to download the image, and in this case Gladys incorrectly displays “no new image” (I will fix this message).
To find the real cause, you can connect via SSH to the machine running Gladys and run:
df -h
docker pull gladysassistant/gladys:v4
If df -h shows a nearly full disk, you can free up space with docker image prune -a (this deletes unused Docker images, not your data), then run the docker pull again.
If the docker pull displays an error (toomanyrequests, no space left on device, network error…), copy it here.
Once the docker pull is successful, you can restart the update from Settings > System. If it still gets stuck, you can do it manually:
docker run --rm -v /var/run/docker.sock:/var/run/docker.sock nickfedor/watchtower:1.20.2 --run-once --cleanup gladys
(replace gladys with the name of your container if needed, visible with docker ps).
Otherwise, if you saw log lines appear under the error message after clicking “Update”, a screenshot would also help.
