/video signaling proxy and a global public server http path config item

docs/server-admin-container-migration.md

@@ -6,7 +6,8 @@ This document is the live implementation tracker for the migration.
 
 - [x] Phase 1, step 1: Establish the single data-directory contract
 - [x] Phase 1, steps 2-5: Configuration database, manual setup-file import, setup, and centralized admin UI
-- [ ] Phase 1, steps 6-9: Restart, backup/restore, and internal video proxy
+- [x] Phase 1, steps 6-8: Restart, backup/restore, and internal video proxy
+- [ ] Phase 1, step 9: Legacy-deployment verification and removal of the external `/video` route
 - [ ] Phase 2: Containerization, GHCR publishing, and container lifecycle controls
 
 The single data-directory implementation and local verification are complete. Real snapshot generation, legacy-directory cleanup, and runtime filesystem tracing remain deployment checks for the actual server; they do not leave the implementation step open.
@@ -24,6 +25,7 @@ Phase 1 must be complete and verified before Phase 2 begins. Containerization mu
 - 2026-09-14: Render the schema-driven configuration editor as a YAML-like tree inside one `CardFrame`. Every object or array introduces an ordered header and one indentation guide, every scalar occupies one key/value row, and array operations remain beside their item instead of moving to the far edge. Keep all route-specific RJSF styling in `webui/src/admin/styles.css`, outside the shared global stylesheet.
 - 2026-09-14: Restart only the application process, never the host. A lockdown administrator with recent password confirmation requests one audited restart, Node acknowledges and announces it, then sends itself SIGTERM. Existing service signal handlers clean up their owned children, while systemd `Restart=always` and the later container restart policy start the application again.
 - 2026-09-14: Keep backup and restore together in one server service after application restart exists. Neither operation stops running services or writers, and no command-line interface is maintained. Backup uses online SQLite snapshots and stable copies of non-database files; restore validates and stages an uploaded archive, records a marker, and uses the normal application restart to replace the data directory during earliest startup.
+- 2026-09-14: Define this server's canonical `publicUrl` once at the top of configuration beside `timezone`. Discord links, inter-instance identity, page metadata, and MediaMTX's primary public ICE hostname derive from it. Browser WHEP and WHIP signaling always uses the same-origin `/video` path; `media.additionalHosts` remains only for genuinely additional ICE names or addresses.
 - 2026-09-14: Treat container deployment as a fresh installation. Neither startup nor the installer searches for, imports, removes, or otherwise manages an old `config.yaml`; the only old-file paths retained are operator-selected YAML uploads on `/setup` and the protected Configuration page. The separate command-line importer and its dry-run mode are removed. Internal SQLite schema migrations remain because they evolve the active database rather than discovering an old installation.
 - 2026-09-14: Keep the one-time first-run setup code in `data/setup-code.txt` with owner-only permissions instead of writing the credential into server logs. Reuse it across restarts and delete it permanently when setup completes.
 - 2026-09-14: Feature enablement is exactly the service-owned `enabled` boolean. A service-owned configuration definition marks itself with `feature: true` when that switch belongs in the public feature map; the configuration system derives the map for sessions and command availability, including nested service definitions, without a separate feature registry. Missing credentials, hardware, connections, data, or enabled dependencies are runtime health conditions and never silently change that choice.
@@ -460,11 +462,13 @@ Implemented on 2026-09-14:
 - Added one protected backup-and-restore service and admin page. Backups keep the application online, use SQLite's online snapshot API for all three databases, make verified stable copies of the remaining durable files, and produce a checksummed archive through a short-lived one-use download. Restore uploads are size-limited, reject unsafe archive entries, verify the complete manifest, checksums, SQLite integrity, and supported schema versions, then remain staged until explicit recent-password confirmation.
 - Restore now uses the normal application restart rather than stopping services itself. The earliest server startup swaps the validated replacement into the data directory, retains one rollback copy, and removes that copy only after the restored application reaches a stabilization point; an interrupted or failed first startup automatically puts the previous data back on the following start. Backup/restore control files and all staging remain inside `data/backup-restore`.
 - Fixed production WAL-mode snapshots creating unmanifested SQLite `-wal` and `-shm` files during schema inspection. Backup and restore validation now remove only those temporary staged sidecars before archiving or applying data, and the regression fixture uses WAL mode to match the real databases.
+- Added the early streaming `/video` middleware with `http-proxy-middleware`. Express removes the public prefix before forwarding WHEP/WHIP requests to `127.0.0.1:8889`, while root-relative MediaMTX session locations receive the prefix again so subsequent browser `PATCH` and `DELETE` requests follow the same path. MediaMTX signaling now binds to loopback; its ICE UDP/TCP listener remains directly reachable on port 8189.
+- Replaced Discord's `siteUrl`, the inter-instance profile's `publicUrl`, and media `whepBaseUrl` with one top-level `publicUrl`. A numbered internal database migration transforms every saved configuration revision before current validation, and the media section now contains only optional additional ICE hosts. WHEP and microphone WHIP URLs are fixed relative paths, so they work through the current origin without knowing its hostname.
 
 Local verification completed:
 
-- All 114 server tests passed, including populated legacy-style default coverage, complete schema-description and input-example coverage, file-backed setup-code lifecycle and symlink rejection, service-definition-derived feature projection, schema-derived secret paths, configuration defaults and strict validation, full-document revision conflicts, secret preservation, administrator invariants, setup and initialized-server YAML import safety, recursive removal of nonexistent fields, and the earlier filesystem coverage.
-- All 25 server test files passed after live application and backup/restore were added. The isolated backup/restore tests cover production-style WAL snapshots without transient sidecars, complete archive round trips, excluded runtime/control data, checksum tampering, unsafe symbolic-link entries, earliest-startup replacement, successful cleanup, and automatic rollback. Application-restart syntax, authorization wiring, and supervisor configuration were checked without exercising the real process signal on the development machine.
+- All 118 server tests passed, including populated legacy-style default coverage, complete schema-description and input-example coverage, file-backed setup-code lifecycle and symlink rejection, service-definition-derived feature projection, schema-derived secret paths, configuration defaults and strict validation, full-document revision conflicts, secret preservation, administrator invariants, setup and initialized-server YAML import safety, recursive removal of nonexistent fields, and the earlier filesystem coverage.
+- All 26 server test files passed after live application, backup/restore, and the internal media proxy were added. The media tests stream exact SDP and trickle-ICE bodies through `POST`, `PATCH`, and `DELETE`, preserve headers, verify prefix and session-location rewriting, confirm loopback-only signaling, and derive the public ICE hostname from the canonical URL. The database migration and production-style WAL backup/restore paths are also covered. Application restart was not signaled on the development machine.
 - Focused admin, route, and identity UI lint passed.
 - All 20 existing focused web UI tests passed.
 - The production web UI build completed successfully and regenerated the checked-in server assets.
@@ -683,7 +687,7 @@ Within the two hard phase boundaries, the safest order is:
 - [x] Apply every configuration revision to running services without restarting the application.
 - [x] Standardize graceful application restart.
 - [x] Implement online backup and restart-bound staged restore in one service.
-- [ ] Add the internal `/video` proxy and remove the special external route.
+- [x] Add the internal `/video` proxy and make the special external route unnecessary.
 - [ ] Run the full Phase 1 completion gate on the legacy deployment.
 - [ ] Build and verify the production application image.
 - [ ] Add Compose, data mounting, networking, and hardware access.