fed307e75c
All checks were successful
PR build (required check) / changes (pull_request) Successful in 7s
secrets-guard / encrypted (pull_request) Successful in 5s
PR build (required check) / build-backend (pull_request) Has been skipped
shell-lint / shellcheck (pull_request) Successful in 6s
PR build (required check) / build-frontend (pull_request) Has been skipped
PR build (required check) / validate-observability (pull_request) Has been skipped
PR build (required check) / gate (pull_request) Successful in 2s
Dev was mesh-only: bound to 10.10.0.2, no DNS record, no Caddy site. A public
record now exists for it, so it gets a real front door instead of a name that
resolves to a timeout.
Three things keep that acceptable on a box that also runs Forgejo and its CI
runner, and all three are load-bearing together:
- the stack binds LOOPBACK, so the site block is the only route in — not even
the mesh reaches the app directly, which is what makes the guard total
rather than decorative,
- HTTP basic auth, so a stumbled-upon URL is not a running build,
- X-Robots-Tag: noindex, so dev never competes with thermograph.org for the
same content in a search index.
The credential in Caddyfile.vps1 is a bcrypt hash, not a password.
|
||
|---|---|---|
| .. | ||
| backup | ||
| db | ||
| forgejo | ||
| migrations | ||
| openmeteo | ||
| secrets | ||
| stack | ||
| swarm | ||
| twa | ||
| Caddyfile.vps1 | ||
| Caddyfile.vps2 | ||
| deploy-dev.sh | ||
| deploy.sh | ||
| env-topology.sh | ||
| migrate-db.py | ||
| POSTGRES-MIGRATION.md | ||
| provision-agent-access.sh | ||
| provision-dev.sh | ||
| provision-mail-supervision.sh | ||
| provision-mail.sh | ||
| provision-secrets.sh | ||
| render-secrets.sh | ||
| RUNBOOK-vps1-vps2-cutover.md | ||
| thermograph-dev.service | ||
| thermograph.env.example | ||
| thermograph.service | ||