fix(rustrak): point templates at the official images and current metadata

[AbianS]

I maintain Rustrak and submitted these two templates a while back. They have gone stale and are handing users a broken install, so this brings them up to date.

What was wrong

Both blueprints pointed at abians7/* images pinned to v0.2.1. The current release is v0.14.2, so anyone deploying from Dokploy was getting a build twelve minor versions old from a personal Docker Hub namespace. The metadata linked to the old personal repo, and rustrak-full also linked to a docs site that has since moved.

What changed

	Before	After
rustrak	abians7/rustrak-server:v0.2.1	rustrak/rustrak-server:latest
rustrak-full server	abians7/rustrak-server:v0.2.1-postgres	rustrak/rustrak-server:postgres
rustrak-full UI	abians7/rustrak-ui:v0.1.6	rustrak/rustrak-ui:latest

• version is now latest, matching the convention other :latest templates use.
• Links point at github.com/rustrak/rustrak and rustrak.github.io/rustrak.
• Logo replaced with the project's current mark.
• The rustrak description carried an estimated <100MB memory footprint. It now says 32 MB at idle, measured with docker stats on the image this template deploys. The rustrak-full description drops the figure entirely, since that stack includes PostgreSQL and I have not measured it.

No changes to template.toml, the compose structure, or any other blueprint.

Tested

Both templates were deployed locally from these exact files before opening this PR.

• rustrak: server starts, migrations run, superuser is created, Actix listens on 8080.
• rustrak-full: PostgreSQL passes its healthcheck, the server connects and migrates against it, the dashboard starts and reaches the API.
• docker compose config valid for both, and node build-scripts/generate-meta.js --check passes on all 517 templates.

[github-actions]

built with Refined Cloudflare Pages Action

⚡ Cloudflare Pages Deployment

Name	Status	Preview	Last Commit
templates	✅ Ready (View Log)	Visit Preview	ee29d99

[Siumauricio]

Tested this PR on a real Dokploy instance. Summary: the rustrak template works out of the box with the new image; rustrak-full needs a one-line addition for the UI to be reachable (pre-existing issue, not caused by this PR).

Upstream / images

• github.com/AbianS/rustrak now 301-redirects to github.com/rustrak/rustrak (org, 105 stars, 18 forks, 742 commits by @AbianS, releases v0.14.3 through v0.14.7 published by @AbianS), so the author is the upstream maintainer and the new links/logo are the current ones.
• Docker Hub manifests all resolve (HTTP 200, anonymous pull): rustrak/rustrak-server:latest, rustrak/rustrak-server:postgres, rustrak/rustrak-ui:latest. The latest/postgres tags were pushed 2026-08-23 together with v0.14.7 / v0.14.7-postgres, matching the latest GitHub release.
• node generate-meta.js --check (517 templates), validate-docker-compose.ts and validate-template.ts pass for both rustrak and rustrak-full.

Deploy: rustrak (SQLite) – works

• Deploy done in 10s, container running, RestartCount 0.
• Logs: Database migrations completed successfully, Superuser created: <email>, starting service: "actix-web-service-0.0.0.0:8080".
• Through Traefik: GET /health -> 200 {"status":"ok"}, GET / -> 401 JSON (auth-protected API, expected).

Deploy: rustrak-full (server + UI + Postgres) – server OK, UI 502

• postgres healthy, rustrak-server (:postgres tag) migrates against it and creates the superuser; API domain answers 401 as expected.
• rustrak-ui domain answers 502 Bad Gateway from Traefik even though the container is running. The container log shows the reason:

  ▲ Next.js 16.3.1
  - Local:   http://5bf9eab913dd:3000
  

  The Next.js standalone server.js binds to $HOSTNAME (the container id inside Docker) instead of 0.0.0.0. On Dokploy the container is attached to two networks (compose network + dokploy-network), so Traefik cannot reach the socket.
• I redeployed the current canary version of this template (abians7/rustrak-ui:v0.1.6) and it fails with the exact same 502, so this is pre-existing and not a regression introduced here — but since we're touching the template anyway it's worth fixing.
• Fix verified on the instance (UI -> 200, page title Rustrak - Error Tracking System, API -> 401):

   rustrak-ui:
     image: rustrak/rustrak-ui:latest
     restart: unless-stopped
     environment:
       - RUSTRAK_API_URL=${RUSTRAK_API_URL}
  +    - HOSTNAME=0.0.0.0

  Could you add that line to blueprints/rustrak-full/docker-compose.yml? (I could not push to your branch from this session.)

Non-blocking notes

• Upstream docker-compose.yml mounts sourcemap_data:/data/sourcemaps on the server (SOURCEMAP_STORAGE_PATH). In rustrak-full the server has no volume, so uploaded sourcemaps would be lost on recreate. Consider adding a rustrak_data:/data volume to rustrak-server in the full stack, like the SQLite template already has.
• :latest/:postgres are acceptable here (many templates in this repo use latest), but if you prefer reproducible deploys the versioned tags v0.14.7, v0.14.7-postgres and rustrak-ui:v0.14.7 exist on Docker Hub.

With the HOSTNAME line added this is good to merge from my side.