fix(relay): stop bystander instances from capturing the relay host

Two mitigations for local multi-instance contention over the shared relay
identity:

- A standby instance now waits a 2-minute grace period after the host claim
  frees before taking over, so a cleanly restarting host (app update or
  relaunch) — which reclaims at boot with no wait — always wins the restart
  window instead of stranding paired devices on another process.
- Dev instances never host the relay passively: dev scripts set
  OPENCHAMBER_RELAY_HOST=off and the Electron dev shell is detected via
  OPENCHAMBER_ELECTRON_DEV. Explicit enable/pairing on such an instance still
  force-claims; OPENCHAMBER_RELAY_HOST=on overrides.

CHANGELOG.md

@@ -18,6 +18,8 @@ All notable changes to this project will be documented in this file.
 - Sessions: headers now find archived sessions too, so an archived session's title no longer goes missing.
 - UI: the timeline dialog now fits small screens instead of squeezing the message list to a couple of rows (thanks to @gaojunran).
 - UI: the btw panel's shadow is lighter, matching the composer.
+- Devices: re-pairing a phone (or logging in again) keeps the device's existing name in Connected Devices instead of resetting it to "OpenChamber Mobile".
+- Relay: paired devices no longer get logged out when the app restarts (for example during an update) while another local OpenChamber process is running — the restarted app keeps serving them instead of a bystander process taking over.
 
 ## [1.20.0] - 2026-08-23
 

packages/web/package.json

@@ -13,8 +13,8 @@
   },
   "scripts": {
     "dev": "bun run build:watch",
-    "dev:server": "bun server/index.js --port ${OPENCHAMBER_PORT:-3001}",
+    "dev:server": "OPENCHAMBER_RELAY_HOST=${OPENCHAMBER_RELAY_HOST:-off} bun server/index.js --port ${OPENCHAMBER_PORT:-3001}",
-    "dev:server:watch": "nodemon --watch server --ext js --exec \"bun server/index.js --port ${OPENCHAMBER_PORT:-3001}\"",
+    "dev:server:watch": "OPENCHAMBER_RELAY_HOST=${OPENCHAMBER_RELAY_HOST:-off} nodemon --watch server --ext js --exec \"bun server/index.js --port ${OPENCHAMBER_PORT:-3001}\"",
     "build": "vite build",
     "build:watch": "vite build --watch",
     "type-check": "tsc --noEmit",

packages/web/server/index.js

@@ -1815,6 +1815,14 @@ async function main(options = {}) {
       fs,
       process,
     }),
+    // Dev/debug instances share the data dir (and thus the relay identity) with
+    // the production instance, so they must not host the relay on their own —
+    // paired devices would land on them. OPENCHAMBER_RELAY_HOST=off disables
+    // passive hosting explicitly (dev scripts set it); the Electron dev shell is
+    // covered via OPENCHAMBER_ELECTRON_DEV. OPENCHAMBER_RELAY_HOST=on overrides
+    // both. Explicit enable/pairing on the instance still hosts regardless.
+    allowPassiveHost: process.env.OPENCHAMBER_RELAY_HOST === 'on'
+      || (process.env.OPENCHAMBER_RELAY_HOST !== 'off' && process.env.OPENCHAMBER_ELECTRON_DEV !== '1'),
     // Relay demand = any paired device or pending pairing session that uses the
     // relay transport. Drives the auto on/off lifecycle.
     hasRelayDemand: async () => {

packages/web/server/lib/relay/DOCUMENTATION.md

@@ -23,7 +23,7 @@ Host side (`packages/web/server/lib/relay/`):
 - `identity.js` — the host's stable identity: the long-lived signing keypair (shared with the push relay, defines the routing id) plus a long-lived encryption keypair (the E2EE trust anchor). Reused across restarts; never rotated implicitly.
 - `signing-key.js` — storage/derivation of the signing keypair and the routing id, shared with the notifications runtime.
 - `host-client.js` — the long-lived connection manager: one outbound control connection to the relay, a per-client data connection for each connected device, reconnect/backoff, and the E2EE responder handshake per connection.
-- `host-lock.js` — the per-machine host claim. Every local instance sharing the data dir shares the relay identity (same serverId), so concurrent relay hosts evict each other at the relay worker (`4001: Control replaced`) and paired devices land on whichever local process won last. The claim file (`<data-dir>/relay-host.lock`, `{ pid }`) makes this deterministic: `service.js` only starts the host when no LIVE process holds the claim (stale claims from dead pids are ignored), goes to `standby` otherwise, and a 30s watcher both takes over when the claimant dies and stands down when another process claims. Explicit user intent — creating a pairing link or hitting `/relay/enable` — force-claims; the previous holder's watcher sees the takeover and backs off instead of fighting. The claim is cooperative (the relay worker still enforces the single host slot); it only decides which process keeps retrying.
+- `host-lock.js` — the per-machine host claim. Every local instance sharing the data dir shares the relay identity (same serverId), so concurrent relay hosts evict each other at the relay worker (`4001: Control replaced`) and paired devices land on whichever local process won last. The claim file (`<data-dir>/relay-host.lock`, `{ pid }`) makes this deterministic: `service.js` only starts the host when no LIVE process holds the claim (stale claims from dead pids are ignored), goes to `standby` otherwise, and a 30s watcher both takes over when the claimant dies and stands down when another process claims. A standby watcher waits a 2-minute grace after the claim frees before taking over, so a cleanly restarting host (app update/relaunch) — which reclaims at boot with no wait — always wins the restart window over a bystander instance. Explicit user intent — creating a pairing link or hitting `/relay/enable` — force-claims; the previous holder's watcher sees the takeover and backs off instead of fighting. Instances created with `allowPassiveHost: false` (dev servers via `OPENCHAMBER_RELAY_HOST=off`, the Electron dev shell via `OPENCHAMBER_ELECTRON_DEV`; `OPENCHAMBER_RELAY_HOST=on` overrides) never start the host passively at all — boot, demand reconcile, and watcher takeover leave them in `standby`; only explicit enable/pairing hosts there. The claim is cooperative (the relay worker still enforces the single host slot); it only decides which process keeps retrying.
 - `tunnel-host.js` — the per-connection dispatcher: decrypts tunnel frames and forwards HTTP/SSE/WS to the local server over loopback, then streams responses back. Enforces a path allowlist and never injects credentials.
 - `e2ee.js`, `tunnel-codec.js` — host-side (JS) mirrors of the shared crypto and framing (see "Two implementations" below).
 

packages/web/server/lib/relay/service.js

@@ -70,6 +70,11 @@ export const createRelayService = ({
   // evict each other at the relay worker ("Control replaced") and devices land
   // on a random instance. Optional: without it, behavior is pre-lock.
   hostLock = null,
+  // When false, this instance never starts the relay host on its own (boot,
+  // demand reconcile, or claim-watch takeover) — only an explicit user action
+  // (enable, pairing) force-claims. Dev/debug instances set this so they do not
+  // capture paired devices from the production instance sharing the data dir.
+  allowPassiveHost = true,
   logger = console,
 }) => {
   const identityRuntime = createRelayIdentityRuntime({ crypto, readSettingsFromDiskMigrated, writeSettingsToDisk, readSettingsStrict });
@@ -80,6 +85,13 @@ export const createRelayService = ({
   // claimant dies; a running host stands down when another process claims.
   let claimWatchTimer = null;
   const CLAIM_WATCH_INTERVAL_MS = 30_000;
+  // A standby instance does not grab a freed claim immediately: a clean restart
+  // of the previous host (app update, relaunch) releases the claim for a short
+  // while, and taking it during that window strands the devices on this —
+  // possibly older — instance. The restarting host reclaims at boot without any
+  // wait, so it always wins the window.
+  const CLAIM_TAKEOVER_GRACE_MS = 120_000;
+  let claimFreeSinceMs = null;
 
   const readConfig = async () => {
     const settings = await readSettingsFromDiskMigrated();
@@ -133,8 +145,19 @@ export const createRelayService = ({
             }
             return;
           }
-          if (status.state === 'standby' && hostLock.tryClaim()) {
+          if (status.state !== 'standby' || !allowPassiveHost) return;
-            logger.warn('[Relay] host claim is free — taking over the relay host');
+          if (hostLock.liveClaimantPid() !== null) {
+            claimFreeSinceMs = null;
+            return;
+          }
+          if (claimFreeSinceMs === null) {
+            claimFreeSinceMs = Date.now();
+            return;
+          }
+          if (Date.now() - claimFreeSinceMs < CLAIM_TAKEOVER_GRACE_MS) return;
+          if (hostLock.tryClaim()) {
+            claimFreeSinceMs = null;
+            logger.warn('[Relay] host claim stayed free — taking over the relay host');
             await start(relayUrl);
           }
         } catch (error) {
@@ -149,10 +172,19 @@ export const createRelayService = ({
     if (!claimWatchTimer) return;
     clearInterval(claimWatchTimer);
     claimWatchTimer = null;
+    claimFreeSinceMs = null;
   };
 
   const start = async (relayUrl, { claim = 'try' } = {}) => {
     if (hostClient) return;
+    if (claim !== 'force' && !allowPassiveHost) {
+      status = {
+        state: 'standby',
+        lastError: 'passive relay hosting is disabled on this instance — enable the relay or create a pairing link to host here',
+        connectedClients: 0,
+      };
+      return;
+    }
     if (hostLock) {
       const claimed = claim === 'force' ? hostLock.forceClaim() : hostLock.tryClaim();
       if (!claimed) {

packages/web/server/lib/relay/service.test.js

@@ -0,0 +1,69 @@
+import { describe, it, expect, vi } from 'vitest';
+import crypto from 'node:crypto';
+
+import { createRelayService } from './service.js';
+
+const makeService = (options = {}) => {
+  // In-memory settings store with a pre-seeded relay identity so the service
+  // never regenerates a signing key during the test.
+  const { privateKey, publicKey } = crypto.generateKeyPairSync('ec', { namedCurve: 'P-256' });
+  let settings = {
+    relaySigningKey: {
+      privateJwk: privateKey.export({ format: 'jwk' }),
+      publicJwk: publicKey.export({ format: 'jwk' }),
+    },
+    privateRelay: { enabled: true, relayUrl: 'wss://relay.example.test/ws' },
+    ...options.settings,
+  };
+  const hostLock = {
+    tryClaim: vi.fn(() => true),
+    forceClaim: vi.fn(() => true),
+    holdsClaim: vi.fn(() => true),
+    liveClaimantPid: vi.fn(() => null),
+    release: vi.fn(),
+  };
+  const service = createRelayService({
+    crypto,
+    readSettingsFromDiskMigrated: async () => settings,
+    writeSettingsToDisk: async (next) => { settings = next; },
+    readSettingsStrict: async () => settings,
+    getLocalPort: () => 0,
+    hasRelayDemand: options.hasRelayDemand ?? (async () => true),
+    hostLock,
+    allowPassiveHost: options.allowPassiveHost,
+    logger: { warn: () => {} },
+  });
+  return { service, hostLock, getSettings: () => settings };
+};
+
+describe('relay service passive hosting', () => {
+  it('never claims or starts the host passively when passive hosting is disabled', async () => {
+    const { service, hostLock } = makeService({ allowPassiveHost: false });
+    try {
+      await service.startIfEnabled();
+      let status = await service.getStatus();
+      expect(status.state).toBe('standby');
+      expect(hostLock.tryClaim).not.toHaveBeenCalled();
+      expect(hostLock.forceClaim).not.toHaveBeenCalled();
+
+      await service.reconcile();
+      status = await service.getStatus();
+      expect(status.state).toBe('standby');
+      expect(status.lastError).toContain('passive relay hosting is disabled');
+      expect(hostLock.tryClaim).not.toHaveBeenCalled();
+    } finally {
+      service.stop();
+    }
+  });
+
+  it('force-claims for an explicit pairing even when passive hosting is disabled', async () => {
+    const { service, hostLock } = makeService({ allowPassiveHost: false });
+    try {
+      const candidate = await service.ensureEnabledForPairing();
+      expect(candidate?.type).toBe('relay');
+      expect(hostLock.forceClaim).toHaveBeenCalled();
+    } finally {
+      service.stop();
+    }
+  });
+});

scripts/dev-web-hmr.mjs

@@ -114,6 +114,9 @@ clearViteCache();
 
 const api = run('api', 'bun', ['run', '--cwd', 'packages/web', 'dev:server:watch'], {
   OPENCHAMBER_PORT: backendPort,
+  // Dev backends share the relay identity with the production instance; never
+  // let them capture the machine's relay host on their own.
+  OPENCHAMBER_RELAY_HOST: process.env.OPENCHAMBER_RELAY_HOST || 'off',
 });
 const vite = run(
   'vite',