Neon LawCall +1 510 800 2090
  • Navigator
  • Notations
  • Legal Aid
  • Sign in

← Operating Neon Law Navigator

Prepare Google Cloud

Chapter 2 of 7 · Section 5 of 54

Sections

1. Intro

  • 1. Deploy your own
  • 2. Agenda

2. Prepare Google Cloud

  • 3. Bring your own project
  • 4. Dry-run first
  • 5. Private assets and domain restricted sharing
  • 6. The Navigator deployment matrix
  • 7. The `/app` mount and HTTP route ownership
  • 8. `neon` — the whole brand seam
  • 9. Live rollout checkpoint
  • 10. Set one site to one version

3. Provision the Infrastructure

  • 11. The APIs that light up
  • 12. Network and storage buckets
  • 13. How a matter's pages reach a client
  • 14. A document's bytes and authorization are separate
  • 15. A private image registry
  • 16. The cluster comes up
  • 17. What setup owns — and what follows it

4. Operate Forgejo beside Navigator

  • 18. Deploy the forge before switching repositories
  • 19. Separate people, Git transport, and machine access
  • 20. Require the gate and protected-main policy
  • 21. Bind the workflow to one Project and one commit
  • 22. Notify failures even when the manifest is broken
  • 23. Restore the repository before cutting over

5. Environment Matrix

  • 24. Three operating modes, two deployment profiles
  • 25. The deployment that says its matters are sample
  • 26. Deployment and release switches
  • 27. Workflow host and routing switches
  • 28. Capability and storage switches
  • 29. Configuration precedence: the first source wins
  • 30. Local dev controls: inputs read by `hudson dev`
  • 31. Local runtime: what `.devx/env` generates
  • 32. The store: SurrealDB
  • 33. Where SurrealDB authorization lives
  • 34. Deployed runtime: core web and worker wiring
  • 35. Deployed runtime: identity and access
  • 36. Deployed runtime: email, signatures, and billing
  • 37. Deployed runtime: repositories, content, AI, and scheduled work
  • 38. Provision and ship: variables read by the operator CLI
  • 39. Ancillary operations and opt-in test controls
  • 40. When sample data appears

6. Configure the Trust Boundaries

  • 41. Secrets: the invariants that gate the boot
  • 42. Secret access, encryption, and audit
  • 43. Sign-in: bring an OIDC provider; passwords live there, not here
  • 44. Role rings: who can do what
  • 45. Provider setup and parity across deployment rows
  • 46. The external surface — every third party, in one place
  • 47. The two service deployments
  • 48. Security architecture

7. Ship the Instance

  • 49. Ship and verify
  • 50. Point your domain at the instance (optional)
  • 51. Drive it from the CLI
  • 52. Make it yours — white-label under your own brand
  • 53. First-ship lessons
  • 54. Canonical references

Private assets and domain restricted sharing

Every deployment bucket remains private. hudson ops gcp setup grants the deployment Google service account roles/storage.objectAdmin on its assets, documents, exports, logs, and applications buckets, then maps both Kubernetes service accounts (navigator-web and workflows-service) to that Google identity through Workload Identity. The PostgreSQL backups bucket is the exception: only the navigator-postgres-backup Kubernetes service account gets object admin on it, and the web identity gets nothing. It never adds allUsers and never changes constraints/iam.allowedPolicyMemberDomains. The applications bucket is still provisioned, but nothing reads it: the Portal is retired, and /app/projects/{code}/portal redirects to the matter show page.

The public website receives only marketing bytes through GET /assets/*. web reads that object from NAVIGATOR_ASSETS_BUCKET with the deployment identity and returns it with its stored content type, X-Content-Type-Options: nosniff, and a one-hour public cache. Unsafe keys and missing objects return 404; storage failures return 502. There is no parallel anonymous route for documents, exports, or logs. This is an application delivery boundary, not a Navigator persons.role or Project-participation grant.

Set NAVIGATOR_ASSET_BASE_URL=$NAV_BASE_URL/assets in each deployment's config.toml — it is a plaintext coordinate the ops ship preflight requires. Setup also gives the active gcloud identity the same bucket-scoped object CRUD role so the operator and the Kubernetes runtime can inspect and repair objects without a public or project-wide grant.

Publish the public asset inventory before ops ship. A full or image-only roll reads NAVIGATOR_ASSETS_BUCKET and stops before any rollout when a required object is missing or the bucket cannot be read. The preflight and the post-roll check use one inventory. On a source checkout that inventory is every markdown image under server/content, every responsive gallery variant, the brand media keys, and every licensed webfont face. On a deploy-only checkout (--deployments-dir with no server/content directory) blog references are absent, and the inventory is the workshop images embedded in the CLI plus those same gallery variants, brand media keys, and fonts. Restart-only skips both checks.


hudson ops assets build --src /path/to/source-jpegs
hudson ops assets upload
hudson ops assets fonts upload --family eb-garamond --dir /path/to/woff2
hudson site asset upload --host "$NAVIGATOR_PUBLIC_HOST" "$ASSET_PATH"

assets build then assets upload publishes the gallery variants. assets fonts upload publishes one licensed family. site asset upload publishes one finished image, including brand media and a workshop slide. Run the publish that covers the inventory, then ops ship.

After setup, verify both identities and reject an anonymous principal:


runtime_gsa="$NAVIGATOR_GCP_SERVICE_ACCOUNT_ID@$NAVIGATOR_GCP_PROJECT_ID.iam.gserviceaccount.com"
operator_account="$(gcloud config get-value account)"
case "$operator_account" in
  *.gserviceaccount.com) operator_member="serviceAccount:$operator_account" ;;
  *) operator_member="user:$operator_account" ;;
esac

for bucket_name in \
  "$NAVIGATOR_ASSETS_BUCKET" \
  "$NAVIGATOR_DOCUMENTS_BUCKET" \
  "$NAVIGATOR_EXPORTS_BUCKET" \
  "$NAVIGATOR_LOGS_BUCKET"
do
  iam_rows="$(
    gcloud storage buckets get-iam-policy "gs://$bucket_name" \
      --flatten='bindings[].members[]' \
      --format='value(bindings.role,bindings.members)'
  )"
  printf '%s\n' "$iam_rows" |
    awk -v runtime="serviceAccount:$runtime_gsa" -v operator="$operator_member" '
      $1 == "roles/storage.objectAdmin" && $2 == runtime { runtime_ok = 1 }
      $1 == "roles/storage.objectAdmin" && $2 == operator { operator_ok = 1 }
      $2 == "allUsers" { public = 1 }
      END { if (!runtime_ok || !operator_ok || public) exit 1 }
    '
done

Presenter notes

No output and a zero exit status prove that both principals have roles/storage.objectAdmin and allUsers is absent. The dogfood driver adds the active operator binding when it is absent; the Rust provisioner owns the runtime binding.


Pause on the trust boundary: the organization policy correctly rejects allUsers, so Navigator never solves asset delivery by making a bucket public. Walk the verification output for both named principals, then contrast the one public /assets/* application route with the absence of anonymous document, export, or log routes. The important result is private object CRUD for the operator and workload identity, not a project-wide storage role.


The certificates are the part worth dwelling on. A classic Google-managed certificate is validated by a CA calling the load balancer, so DNS has to point there first — moving a live hostname takes it down for the length of issuance, and Cloud CDN and the HTTP-to-HTTPS redirect both sit in the validation path. Certificate Manager with a DNS authorization avoids that: Google returns a CNAME, the CA reads that record, and the certificate reaches ACTIVE while the hostname still serves its current site. Publish the CNAME first, wait for ACTIVE, then move the A record — a cutover done that way carries no TLS gap. The GKE ingress here owns its own certificates, so it is the classic path above rather than this one.


Name a hostname collision before a cutover. A hostname, certificate, and authorized OAuth redirect URI must belong to one deployment at a time. When moving a hostname, stage its new certificate and routing configuration before changing DNS, and retire the prior route only after the new endpoint is verified.

View all slidesOpen display
← PreviousNext →
Neon Law
  • X opens in a new tab
  • LinkedIn opens in a new tab
  • YouTube opens in a new tab
  • API
  • Blog
  • Contact
  • Design
  • Glossary
  • Navigator
  • Notations
  • Presentations
  • Privacy
  • Team opens in a new tab
  • Terms
  • Testimonials
  • contact@neonlaw.com
  • +1 510 800 2090
  • California
    1515 20th StreetBakersfield, CA 93301
  • Nevada
    5150 Mae Anne AveSte 405-9002Reno, NV 89523
  • New York
    12 E 49th St18th FloorNew York, NY 10017
  • Washington
    1111 Main StSTE 201EVancouver, WA 98660
  • Justice Technology AssociationMission-Aligned Partner opens in a new tab
  • Proud member of Greater Vancouver ChamberGVC opens in a new tab

Attorney advertisement. Nothing here is legal advice without a signed retainer for an active project. Past results do not guarantee future outcomes.

Everyone deserves access to justice.Made with ❤️ in 🗽.

© 2026 Guieb & Shook PC

Powered by neon-law/navigator opens in a new tab26.11.17-rc.1GitHub stars 5