HACS not acknowledging update

[tbclark3]

System Health details

System Information

version	core-2024.8.3
installation_type	Home Assistant Container
dev	false
hassio	false
docker	true
user	root
virtualenv	false
python_version	3.12.4
os_name	Linux
os_version	6.10.6-200.fc40.x86_64
arch	x86_64
timezone	America/Los_Angeles
config_dir	/config

Home Assistant Community Store

GitHub API	ok
GitHub Content	ok
GitHub Web	ok
HACS Data	ok
GitHub API Calls Remaining	4992
Installed Version	2.0.0
Stage	running
Available Repositories	1399
Downloaded Repositories	33

Home Assistant Cloud

logged_in	false
can_reach_cert_server	ok
can_reach_cloud_auth	ok
can_reach_cloud	ok

Dashboards

dashboards	16
resources	17
views	15
mode	storage

Recorder

oldest_recorder_run	August 16, 2024 at 4:18 PM
current_recorder_run	August 25, 2024 at 2:49 PM
estimated_db_size	1555.95 MiB
database_engine	sqlite
database_version	3.45.3

Checklist

• [X] I'm running the newest version of HACS https://github.com/hacs/integration/releases/latest
• [X] I have enabled debug logging for my installation.
• [X] I have filled out the issue template to the best of my ability.
• [X] I have read https://hacs.xyz/docs/help/issues/
• [X] This issue is related to the backend (integration part) of HACS.
• [X] This issue only contains 1 issue (if you have multiple issues, open one issue for each issue).
• [X] This is a bug and not a feature request.
• [X] This issue is not a duplicate issue of currently open or issues pending release.

Describe the issue

After updating an integration and restarting, HACS returns to the "update needed" for that integration even though it is clear that the new version is running. This initially happened with a single integration and remained in the "update needed" state after I updated and restarted several times. I then got a notice of a second integration needing update, proceeded, restarted, and both showed successful updates. A few minutes ago I got another upgrade notice. I updated, restarted, and now HACS is showing "update needed" for that integration. I have confirmed in the integration's manifest that the newest version was downloaded and is running.

Reproduction steps

1. Update integration
2. Check HACS page
3. HACS still says integration needed ...

Debug logs

Logs don't show anything significant

Diagnostics dump

No response

[hacs-bot[bot]]

Make sure you have read the issue guidelines and that you filled out the entire template.

If you have an issue identical to this, do not add comments like "same here", "i have this too", instead add a :+1: reaction to the issue description. Thanks! :+1:

[wizu]

I have same problem, it looks like hacs only see intergration as updated, when I aceept notification about restart and it automatically restarts HA. After manual restart, notification persists.

[tbclark3]

Good pickup. That explains why I see it every time. I can't restart HA from within HA because I have to restart an stunnel instance at the same time, so I always restart using docker.

[ludeeus]

HACS flushes the metadata to disk every 6h and on the shutdown event (which also happens during restart). You need to allow more time before killing it if you are not using the UI to restart. Closing as out of scope for HACS.

[wizu]

I use docker compose with stop_grace_period: 60s. When i do "docker compose stop homeassistnat" then docker sends SIGTERM to container and waits 60s - after that time it sends SIGKILL. HA shut down in about 10s so no SIGKILL is send. That clearly shows that HA completes it's shutdown procedure and it is not interrupted. So HACS should also have time to do what it is supposed to do - but it clearly don't.

[wizu]

Oh, it may be the case that my new value for stop_grace_period was not applied correctly. I will observe it with next restarts and leave feedback here.

[wizu]

@ludeeus @tbclark3 stop_grace_period: 60s did the trick. Now this notification don't show up after restart through docker.