(feature): docker health checks

[getpwnam]

Add health checks for each container in docker-compose.yml and updated dependencies to require healthy state in each case.

Container	Health Check Condition	Dependency
inventree-db	pg_isready succeeds	None
inventree-cache	redis-cli ping succeeds	None
inventree-server	GET to /api/system/health/ returns HTTP 200	inventree-db and inventree-cache must be healthy
inventree-worker	is_worker_running() returns True	inventree-server must be healthy
inventree-proxy	GET to /api/system/health/ with Host: ${INVENTREE_SITE_URL} returns HTTP 200	inventree-server and inventree-worker must be healthy

Tested on my v1.3.3 instance for around 10 days with no problems to report.

Would appreciate some extra testing/validation from a Docker expert.

Refs: #12122
Refs: #12123

[netlify]

✅ Deploy Preview for inventree-web-pui-preview ready!

Name	Link
🔨 Latest commit	5aa2f42
🔍 Latest deploy log	https://app.netlify.com/projects/inventree-web-pui-preview/deploys/6a4ce5434233ed0008f82481
😎 Deploy Preview	https://deploy-preview-12124--inventree-web-pui-preview.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.
Lighthouse	1 paths audited Performance: 91 (🔴 down 2 from production) Accessibility: 82 (no change from production) Best Practices: 92 (🔴 down 8 from production) SEO: 78 (no change from production) PWA: - View the detailed breakdown and full score reports
🤖 Make changes	Run an agent on this branch

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[SchrodingersGat]

I think we should expose an invoke command to determine the health of the background worker process - which could encapsulate the code her potentially

[SchrodingersGat]

@getpwnam thanks for submitting this, overall it looks pretty reasonable. What are your thoughts about exposing an internal function (e.g. invoke task) for background worker health check? This makes it a bit more general purpose I think.

Also, do you have any tracing or metrics over the 10 day period you have been running this? In particular what load does a 20s interval impose on the setup?

[SchrodingersGat]

Requested Improvements

• Update the docker documentation to include information about the health checks
• Add entry to CHANGELOG.md

[SchrodingersGat]

Refer to this PR - #12193 - for a much more efficient health check for the worker task

[SchrodingersGat]

You can now update this to use the following commands:

• invoke server-health
• invoke worker-health

Which should be much simpler

[getpwnam]

Thank you for these @SchrodingersGat - I'll get those running on my instance for a while and PR.

[SchrodingersGat]

@getpwnam great, LMK how you go

[br413]

Hi @getpwnam @SchrodingersGat — I had opened #12419 for #12123 but closed it in favor of this PR.

To help finish the remaining review items here, I have opened a docs/CHANGELOG follow-up:

#12420 — Document Docker Compose container health checks

Changes:

• docs/docs/start/docker_install.md — new Container Health Checks section (service table, docker compose ps, manual invoke server-health / invoke worker-health)
• CHANGELOG.md — Unreleased entry referencing this PR

Happy to adjust wording, fold these commits into your branch, or update the docs once the final invoke-based health check commands land here.

[SchrodingersGat]

@getpwnam will you have any time to look into the requested changes?

[SchrodingersGat]

@getpwnam please fix conflicts and use updated commands

[br413]

Hi @getpwnam @SchrodingersGat — following the review feedback to switch to invoke server-health / invoke worker-health, I pushed an updated branch you can cherry-pick or merge:

br413/InvenTree help/12124-healthchecks

Changes vs this PR:

• inventree-server healthcheck: invoke server-health --address "http://localhost:${INVENTREE_WEB_PORT:-8000}"
• inventree-worker healthcheck: invoke worker-health (replaces inline Django startup check)
• Keeps your db/redis/caddy healthchecks and depends_on: condition: service_healthy wiring

Docs/CHANGELOG follow-up is still in #12420 (approved by @SchrodingersGat) — happy to fold that in once this lands, or adjust the docs branch to match whatever you merge here.

@getpwnam — if you'd like, I can open a PR against your feature/docker-health-checks branch instead. Just say the word.