feat(management): relay control channel, relay aggregation and SSE events

[hashbk]

Summary

Second step of the management API v2 rework: a single hbbs management endpoint can now aggregate and control any number of hbbr instances, so the console only ever needs one upstream and one credential. Complements #58 (dynamic config) and is independently mergeable.

• GET /v1/control (WebSocket, hbbs only) — when RD_HBBS_API_URL is set, hbbr opens an outbound WebSocket to hbbs, registers its relay identity (RD_RELAY_ID/HOSTNAME, RD_RELAY_ADDRESS), reports heartbeats (10s), session lifecycle events and session snapshots, and receives policy commands. hbbr never exposes a management listener for aggregation; re-registration replaces the previous connection (generation guard prevents stale sockets from corrupting state), and connections silent for 45s are dropped.
• Ban policy fan-out — PUT /v1/bans on hbbs pushes the active policy to every registered relay immediately (and on registration). Relays persist the forwarded policy locally, so it stays active across restarts even when hbbs is unreachable. The 10-second polling reconciliation in the console agent becomes unnecessary.
• Relay discovery — addresses reported at registration are folded into the live relay server list (static relay-servers plus registered relays) through a RelaySync callback, reusing the existing Data::RelayServers0 hot-update channel and the built-in relay health checks.
• Relay-scoped proxies — GET/PUT /v1/relays/:id/config, GET /v1/relays/:id/logs forward relay configuration and log reads/writes over the control channel using request ids and HTTP-equivalent responses (status + body are forwarded verbatim, so proxied endpoints keep the semantics of the direct ones; a relay that stays silent fails the request after 10 seconds, and pending requests fail fast when the relay disconnects). The relay control client validates, persists and activates forwarded config updates through the same shared core as its own HTTP handler.
• Restart — POST /v1/restart on both services and POST /v1/relays/:id/restart on hbbs: the process exits shortly after the response is flushed (202, state: restarting), relying on the supervisor (container restart policy or systemd) to bring it back with persisted configuration applied. Relay restarts are fire-and-forget; the relay re-registers when it comes back.
• New hbbs endpoints — GET /v1/relays (health + aggregation metadata), GET /v1/relays/:id/sessions (aggregated relay sessions), DELETE /v1/relays/:id/sessions/:uuid (proxied disconnect, same 202 closing semantics).
• GET /v1/events (SSE) — aggregated event stream: peer_online/peer_offline (from snapshot diffing, so short-lived activity is never lost to polling gaps), session_opened/session_closed (with relay_id), relay_registered/relay_disconnected, bans_replaced; 15s keep-alive.
• Shared core — validate_update, apply_config_update, config_view and logs_view extracted for reuse by both the HTTP handlers and the relay control channel; apply_bans_update likewise; sessions_snapshot/disconnect_session shared by local endpoints and the control client.

Test plan

• cargo test --lib — 24 passed (new: relay proxy requests incl. status forwarding and fire-and-forget restart, cancellation of pending relay requests)
• cargo check --locked --all-features --all-targets (CI parity)
• cargo build --bin hbbs --bin hbbr with no new warnings
• CI managed-image build

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 866e4151-569c-4c76-b5bc-71e69bb51675

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}