diff --git a/backend/.env.example b/backend/.env.example
index ff03217..612ddb9 100644
--- a/backend/.env.example
+++ b/backend/.env.example
@@ -41,6 +41,25 @@ ADMIN_PASSWORD=ChangeMe123!
 # Path to Firebase service-account JSON (falls back to backend/firebase-service-account.json)
 FIREBASE_SERVICE_ACCOUNT_PATH=
 
+# --- Valkey availability mirror cadences ---
+#
+# All env-driven per backend/CLAUDE.md Config-Source Convention. Defaults match
+# requirement/valkey-online-mirror-plan.md. Floor-clamped by their getters.
+
+# How often the auto-offline sweep checks heartbeat freshness (seconds).
+# The staleness threshold itself (`stale_after_seconds`) is CC-tunable via app_config.
+MITRA_AUTO_OFFLINE_SWEEP_SECONDS=30
+
+# How often heartbeat timestamps are batched from Valkey → Postgres (seconds).
+# Per-ping heartbeat writes go to Valkey only; this preserves forensic
+# `last_heartbeat_at` in Postgres with up to <interval> seconds of lag.
+HEARTBEAT_MIRROR_INTERVAL_SECONDS=60
+
+# How often Valkey state is re-derived from Postgres to heal drift (seconds).
+# Belt-and-braces against failed best-effort Valkey writes, out-of-band Postgres
+# mutations, or evictions. Set to 0 to disable (not recommended).
+VALKEY_ONLINE_MIRROR_SWEEP_SECONDS=300
+
 # --- Phase 5: Xendit (dev-safe defaults: integration disabled) ---
 #
 # Flip XENDIT_ENABLED=true in staging/prod once secret + webhook token are populated.
diff --git a/backend/src/plugins/valkey.js b/backend/src/plugins/valkey.js
index f4885e2..9cb9cc1 100644
--- a/backend/src/plugins/valkey.js
+++ b/backend/src/plugins/valkey.js
@@ -4,10 +4,31 @@ let pub
 let sub
 let client
 
+// 'ready' listeners (registered before connect; fire on initial connect AND each
+// reconnect). Used by services that need to reseed Valkey state from Postgres.
+const readyListeners = new Set()
+
+const attachReadyHandler = (instance) => {
+  instance.on('ready', () => {
+    for (const fn of readyListeners) {
+      // Fire-and-forget; each listener owns its own error handling.
+      Promise.resolve()
+        .then(() => fn())
+        .catch((err) => console.error('[valkey] ready listener failed:', err))
+    }
+  })
+}
+
+export const onValkeyReady = (fn) => {
+  readyListeners.add(fn)
+  return () => readyListeners.delete(fn)
+}
+
 export const getValkeyClient = () => {
   if (!client) {
     const url = process.env.VALKEY_URL || 'redis://localhost:6379'
     client = new Redis(url)
+    attachReadyHandler(client)
   }
   return client
 }
@@ -51,3 +72,25 @@ export const subscribe = (channel, callback) => {
     console.log(`[valkey] unsubscribed from ${channel}`)
   }
 }
+
+// --- Thin wrappers used by the mitra-availability mirror ---
+//
+// Each wrapper uses the shared `client` (separate ioredis instance from pub/sub
+// to keep subscribe state isolated). Callers in services/* wrap these in
+// try/catch and fall back to Postgres on error — see the plan doc.
+
+export const sadd = (key, ...members) => getValkeyClient().sadd(key, ...members)
+export const srem = (key, ...members) => getValkeyClient().srem(key, ...members)
+export const sismember = async (key, member) =>
+  (await getValkeyClient().sismember(key, member)) === 1
+export const smembers = (key) => getValkeyClient().smembers(key)
+export const sdiff = (...keys) => getValkeyClient().sdiff(...keys)
+export const scard = (key) => getValkeyClient().scard(key)
+export const set = (key, value) => getValkeyClient().set(key, value)
+export const get = (key) => getValkeyClient().get(key)
+export const del = (...keys) => getValkeyClient().del(...keys)
+export const incr = (key) => getValkeyClient().incr(key)
+export const decr = (key) => getValkeyClient().decr(key)
+export const exists = (key) => getValkeyClient().exists(key)
+export const pipeline = () => getValkeyClient().pipeline()
+export const multi = () => getValkeyClient().multi()
diff --git a/backend/src/routes/public/mitra.status.routes.js b/backend/src/routes/public/mitra.status.routes.js
index 22038b3..25be5cc 100644
--- a/backend/src/routes/public/mitra.status.routes.js
+++ b/backend/src/routes/public/mitra.status.routes.js
@@ -1,6 +1,8 @@
 import { authenticate } from '../../plugins/auth.js'
 import { getMitraById } from '../../services/mitra.service.js'
 import * as mitraStatusService from '../../services/mitra-status.service.js'
+import * as valkey from '../../plugins/valkey.js'
+import { VK_MITRAS_DEACTIVATED } from '../../services/mitra-status.service.js'
 import { UserType } from '../../constants.js'
 
 export const mitraStatusRoutes = async (app) => {
@@ -27,6 +29,32 @@ export const mitraStatusRoutes = async (app) => {
     request.mitra = mitra
   }
 
+  // Lightweight heartbeat guard: no Postgres SELECT in the hot path. Checks
+  // `mitras:deactivated` in Valkey (maintained on every updateMitraStatus) and
+  // falls back to the full resolveMitra/DB check if Valkey is unreachable so a
+  // Valkey outage doesn't accept heartbeats from deactivated mitras silently.
+  const heartbeatGuard = async (request, reply) => {
+    if (request.auth?.userType !== UserType.MITRA) {
+      return reply.code(403).send({
+        success: false,
+        error: { code: 'FORBIDDEN', message: 'Mitra account required' },
+      })
+    }
+    try {
+      const deactivated = await valkey.sismember(VK_MITRAS_DEACTIVATED, request.auth.userId)
+      if (deactivated) {
+        return reply.code(403).send({
+          success: false,
+          error: { code: 'ACCOUNT_INACTIVE', message: 'Account is inactive' },
+        })
+      }
+      return
+    } catch (err) {
+      console.warn('[heartbeat] valkey check failed, falling back to DB:', err.message)
+      return resolveMitra(request, reply)
+    }
+  }
+
   app.post('/online', { preHandler: [authenticate, resolveMitra] }, async (request, reply) => {
     await mitraStatusService.setOnline(request.mitra.id)
     return reply.send({ success: true, data: { is_online: true } })
@@ -37,8 +65,8 @@ export const mitraStatusRoutes = async (app) => {
     return reply.send({ success: true, data: { is_online: false } })
   })
 
-  app.post('/heartbeat', { preHandler: [authenticate, resolveMitra] }, async (request, reply) => {
-    await mitraStatusService.heartbeat(request.mitra.id)
+  app.post('/heartbeat', { preHandler: [authenticate, heartbeatGuard] }, async (request, reply) => {
+    await mitraStatusService.heartbeat(request.auth.userId)
     return reply.send({ success: true })
   })
 
diff --git a/backend/src/server.js b/backend/src/server.js
index 40d2707..46f583c 100644
--- a/backend/src/server.js
+++ b/backend/src/server.js
@@ -1,7 +1,13 @@
 import 'dotenv/config'
+
 import { buildPublicApp } from './app.public.js'
 import { buildInternalApp } from './app.internal.js'
-import { autoOfflineStaleMitras } from './services/mitra-status.service.js'
+import { autoOfflineStaleMitras, seedFromPostgres, mirrorHeartbeatsToPostgres } from './services/mitra-status.service.js'
+import {
+  getMitraAutoOfflineSweepSeconds,
+  getHeartbeatMirrorIntervalSeconds,
+  getValkeyOnlineMirrorSweepSeconds,
+} from './services/config.service.js'
 import { initFirebase } from './plugins/firebase.js'
 import { restoreActiveTimers } from './services/session-timer.service.js'
 import { expireStalePaymentRequests, registerPairingSubscriber } from './services/payment.service.js'
@@ -23,18 +29,22 @@ const start = async () => {
   }
 
   initFirebase()
+
   const publicApp = await buildPublicApp()
   const internalApp = await buildInternalApp()
 
+  // restoreActiveTimers runs bulk UPDATEs on chat_sessions to clean up stale
+  // ACTIVE/CLOSING rows from before the restart. Run it BEFORE seedFromPostgres
+  // so the seed sees the post-cleanup state and capacity counters are accurate.
+  await restoreActiveTimers()
+  await seedFromPostgres()
+
   await publicApp.listen({ port: PUBLIC_PORT, host: '0.0.0.0' })
   console.log(`Public API listening on port ${PUBLIC_PORT}`)
 
   await internalApp.listen({ port: INTERNAL_PORT, host: INTERNAL_HOST })
   console.log(`Internal API listening on ${INTERNAL_HOST}:${INTERNAL_PORT}`)
 
-  // Restore session timers for active sessions (on server restart)
-  await restoreActiveTimers()
-
   // Phase 5: wire pairing service as a subscriber to payment_request.confirmed events.
   // Must happen AFTER all services are loaded so the subscriber registration sees
   // the EventEmitter set up by payment.service.js at module-load time.
@@ -53,7 +63,8 @@ const start = async () => {
     console.error('Startup reconciliation failed:', err)
   }
 
-  // Auto-offline mitras with stale heartbeat (every 30s)
+  // Auto-offline mitras with stale heartbeat (env-driven cadence, default 30s).
+  // Valkey-driven per requirement/valkey-online-mirror-plan.md.
   setInterval(async () => {
     try {
       const count = await autoOfflineStaleMitras()
@@ -61,7 +72,32 @@ const start = async () => {
     } catch (err) {
       console.error('Auto-offline check failed:', err)
     }
-  }, 30_000)
+  }, getMitraAutoOfflineSweepSeconds() * 1000)
+
+  // Batched heartbeat mirror: Valkey heartbeat timestamps → Postgres
+  // last_heartbeat_at (default 60s). Keeps forensic column current without
+  // per-ping DB writes. One UNNEST UPDATE per tick; idempotent across instances.
+  setInterval(async () => {
+    try {
+      await mirrorHeartbeatsToPostgres()
+    } catch (err) {
+      console.error('Heartbeat mirror failed:', err)
+    }
+  }, getHeartbeatMirrorIntervalSeconds() * 1000)
+
+  // Reconciliation sweep: heal Valkey/Postgres drift (default 300s; 0 disables).
+  // Belt-and-braces against failed best-effort Valkey writes, out-of-band
+  // Postgres mutations, evictions. Idempotent — just runs the seed.
+  const reconciliationSeconds = getValkeyOnlineMirrorSweepSeconds()
+  if (reconciliationSeconds > 0) {
+    setInterval(async () => {
+      try {
+        await seedFromPostgres()
+      } catch (err) {
+        console.error('Valkey reconciliation sweep failed:', err)
+      }
+    }, reconciliationSeconds * 1000)
+  }
 
   // Expire stale payment_requests + reconcile lost subscriber work (every 60s).
   // Pending past expires_at → expired (no failure row).
diff --git a/backend/src/services/closure.service.js b/backend/src/services/closure.service.js
index 8ac574f..55d10b0 100644
--- a/backend/src/services/closure.service.js
+++ b/backend/src/services/closure.service.js
@@ -3,6 +3,7 @@ import { publish } from '../plugins/valkey.js'
 import { clearSessionTimer, clearClosureGraceTimer, startClosureGraceTimer } from './session-timer.service.js'
 import { sendToSessionParticipant } from '../plugins/websocket.js'
 import { sendPushNotification } from './notification.service.js'
+import { recomputeCapacityForMitra } from './mitra-status.service.js'
 import { UserType, SessionStatus, EndedBy, WsMessage } from '../constants.js'
 
 const sql = getDb()
@@ -59,6 +60,7 @@ export const completeSession = async (sessionId) => {
     RETURNING id, customer_id, mitra_id, status, ended_at
   `
   if (!session) return null
+  await recomputeCapacityForMitra(session.mitra_id)
 
   // Notify both parties, FCM fallback if WebSocket is down
   const data = { type: WsMessage.SESSION_COMPLETED, session_id: sessionId }
@@ -109,6 +111,7 @@ export const initiateEarlyEnd = async (sessionId, userType) => {
       code: 'SESSION_NOT_ACTIVE', statusCode: 409,
     })
   }
+  await recomputeCapacityForMitra(session.mitra_id)
 
   clearSessionTimer(sessionId)
   startClosureGraceTimer(sessionId)
diff --git a/backend/src/services/config.service.js b/backend/src/services/config.service.js
index ae1fcbe..fa6e7a3 100644
--- a/backend/src/services/config.service.js
+++ b/backend/src/services/config.service.js
@@ -149,6 +149,34 @@ export const getMitraHeartbeatCadenceSeconds = () => {
   return Number.isFinite(parsed) && parsed >= 5 ? parsed : 30
 }
 
+// --- Valkey availability mirror — env-driven cadences ---
+//
+// Per requirement/valkey-online-mirror-plan.md. All three are operational
+// knobs (env, per backend/CLAUDE.md Config-Source Convention), not
+// operator-tunable. Defaults match the plan; values are floor-clamped.
+
+export const getMitraAutoOfflineSweepSeconds = () => {
+  const raw = process.env.MITRA_AUTO_OFFLINE_SWEEP_SECONDS
+  if (!raw || raw.trim() === '') return 30
+  const parsed = Number.parseInt(raw, 10)
+  return Number.isFinite(parsed) && parsed >= 5 ? parsed : 30
+}
+
+export const getHeartbeatMirrorIntervalSeconds = () => {
+  const raw = process.env.HEARTBEAT_MIRROR_INTERVAL_SECONDS
+  if (!raw || raw.trim() === '') return 60
+  const parsed = Number.parseInt(raw, 10)
+  return Number.isFinite(parsed) && parsed >= 10 ? parsed : 60
+}
+
+export const getValkeyOnlineMirrorSweepSeconds = () => {
+  const raw = process.env.VALKEY_ONLINE_MIRROR_SWEEP_SECONDS
+  if (!raw || raw.trim() === '') return 300
+  const parsed = Number.parseInt(raw, 10)
+  if (parsed === 0) return 0  // explicit disable
+  return Number.isFinite(parsed) && parsed >= 30 ? parsed : 300
+}
+
 // --- Phase 5: Xendit integration ---
 //
 // Env-driven (per backend/CLAUDE.md Config-Source Convention). All five values
diff --git a/backend/src/services/dashboard.service.js b/backend/src/services/dashboard.service.js
index a4e1f44..5147ab0 100644
--- a/backend/src/services/dashboard.service.js
+++ b/backend/src/services/dashboard.service.js
@@ -1,19 +1,33 @@
 import { getDb } from '../db/client.js'
+import * as valkey from '../plugins/valkey.js'
+import { VK_MITRAS_ONLINE } from './mitra-status.service.js'
 import { SessionStatus, TopicSensitivity } from '../constants.js'
 
 const sql = getDb()
 
+// Valkey-fast SCARD with Postgres fallback. The CC dashboard polls every few
+// seconds; SCARD is sub-ms so this keeps the dashboard responsive at any scale.
+const getOnlineMitrasCount = async () => {
+  try {
+    return await valkey.scard(VK_MITRAS_ONLINE)
+  } catch (err) {
+    console.warn('[dashboard] valkey unavailable, falling back to DB:', err.message)
+    const [{ c }] = await sql`SELECT COUNT(*)::int AS c FROM mitra_online_status WHERE is_online = true`
+    return c
+  }
+}
+
 export const getDashboardStats = async () => {
   const [
     [{ active_chats }],
-    [{ online_mitras }],
+    online_mitras,
     [{ pending_requests }],
     [{ sensitive_total }],
     [{ sensitive_last_30d_total }],
     [{ sensitive_last_30d_sensitive }],
   ] = await Promise.all([
     sql`SELECT COUNT(*) AS active_chats FROM chat_sessions WHERE status IN (${SessionStatus.ACTIVE}, ${SessionStatus.PENDING_PAYMENT})`,
-    sql`SELECT COUNT(*) AS online_mitras FROM mitra_online_status WHERE is_online = true`,
+    getOnlineMitrasCount(),
     sql`SELECT COUNT(*) AS pending_requests FROM chat_sessions WHERE status IN (${SessionStatus.SEARCHING}, ${SessionStatus.PENDING_ACCEPTANCE})`,
     sql`SELECT COUNT(*) AS sensitive_total FROM chat_sessions WHERE topic_sensitivity = ${TopicSensitivity.SENSITIVE}`,
     sql`SELECT COUNT(*) AS sensitive_last_30d_total FROM chat_sessions WHERE created_at >= NOW() - INTERVAL '30 days'`,
diff --git a/backend/src/services/extension.service.js b/backend/src/services/extension.service.js
index b02c299..76bba70 100644
--- a/backend/src/services/extension.service.js
+++ b/backend/src/services/extension.service.js
@@ -1,7 +1,7 @@
 import { getDb } from '../db/client.js'
 import { sendToSessionParticipant, isUserOnlineWs } from '../plugins/websocket.js'
 import { extendSessionTimer, clearClosureGraceTimer, startClosureGraceTimer } from './session-timer.service.js'
-import { isMitraReachable } from './mitra-status.service.js'
+import { isMitraReachable, recomputeCapacityBySession } from './mitra-status.service.js'
 import { consumePaymentSession, failPaymentSession, getPaymentSession } from './payment.service.js'
 import { sendPushNotification } from './notification.service.js'
 import {
@@ -104,6 +104,7 @@ export const requestExtension = async (sessionId, customerId, { duration_minutes
 
   // Pause the session
   await sql`UPDATE chat_sessions SET status = ${SessionStatus.EXTENDING} WHERE id = ${sessionId}`
+  await recomputeCapacityBySession(sessionId)
 
   // Resolve timeout once so we can both surface it in the WS payload and start the server-side timer.
   const timeoutMs = await getExtensionTimeoutMs()
@@ -213,6 +214,7 @@ const finalizeExtension = async (extensionId, sessionId, accepted, viaTimeout) =
 
     // Resume session
     await sql`UPDATE chat_sessions SET status = ${SessionStatus.ACTIVE} WHERE id = ${extension.session_id}`
+    await recomputeCapacityBySession(extension.session_id)
 
     // Record transaction
     await sql`
@@ -249,6 +251,7 @@ const finalizeExtension = async (extensionId, sessionId, accepted, viaTimeout) =
     }
 
     await sql`UPDATE chat_sessions SET status = ${SessionStatus.CLOSING} WHERE id = ${extension.session_id}`
+    await recomputeCapacityBySession(extension.session_id)
 
     sendToSessionParticipant(sessionId, UserType.CUSTOMER, {
       type: WsMessage.EXTENSION_RESPONSE,
@@ -331,6 +334,7 @@ const timeoutExtension = async (extensionId, sessionId, mitraId) => {
 
   // Move session to closing & notify both parties (matches the explicit-reject UX).
   await sql`UPDATE chat_sessions SET status = ${SessionStatus.CLOSING} WHERE id = ${sessionId}`
+  await recomputeCapacityBySession(sessionId)
   sendToSessionParticipant(sessionId, UserType.CUSTOMER, {
     type: WsMessage.EXTENSION_RESPONSE,
     accepted: false,
diff --git a/backend/src/services/mitra-status.service.js b/backend/src/services/mitra-status.service.js
index 558deda..5ea6eac 100644
--- a/backend/src/services/mitra-status.service.js
+++ b/backend/src/services/mitra-status.service.js
@@ -1,24 +1,80 @@
 import { getDb } from '../db/client.js'
 import { SessionStatus } from '../constants.js'
 import { getMitraPingConfig, getMaxCustomersPerMitra } from './config.service.js'
-import { subscribe } from '../plugins/valkey.js'
+import * as valkey from '../plugins/valkey.js'
+import { subscribe, onValkeyReady } from '../plugins/valkey.js'
 
 const sql = getDb()
 
-// --- Short-TTL availability cache for the 5s-poll endpoint ---
-// In-memory snapshot { available, count, expiresAt }. The cache:
-//   - is recomputed at most once per AVAILABILITY_TTL_MS (10s backstop)
-//   - is invalidated explicitly when CC changes max_customers_per_mitra (call invalidateAvailabilityCache())
-// This keeps customer polls off the DB hot path while staying close to real time.
-const AVAILABILITY_TTL_MS = 10_000
-let availabilityCache = null // { available, count, expiresAt }
+// Per requirement/valkey-online-mirror-plan.md § Schema.
+export const VK_MITRAS_ONLINE = 'mitras:online'
+export const VK_MITRAS_DEACTIVATED = 'mitras:deactivated'
+export const vkCapacityKey = (mitraId) => `mitra:capacity:${mitraId}`
+export const vkHeartbeatKey = (mitraId) => `mitra:heartbeat:${mitraId}`
 
-export const invalidateAvailabilityCache = () => {
-  availabilityCache = null
+// Rebuilds Valkey availability state from Postgres. Called on backend startup,
+// on Valkey reconnect (via onValkeyReady), and by the reconciliation sweep.
+// Idempotent — DEL + bulk SADD/SET produces the same final state on every run.
+export const seedFromPostgres = async () => {
+  try {
+    const [onlineRows, deactRows, capacityRows] = await Promise.all([
+      sql`SELECT mitra_id FROM mitra_online_status WHERE is_online = true`,
+      sql`SELECT id FROM mitras WHERE is_active = false`,
+      sql`
+        SELECT mitra_id, COUNT(*)::int AS c FROM chat_sessions
+        WHERE mitra_id IS NOT NULL
+          AND status IN (${SessionStatus.ACTIVE}, ${SessionStatus.PENDING_PAYMENT})
+        GROUP BY mitra_id
+      `,
+    ])
+
+    const pipe = valkey.pipeline()
+    pipe.del(VK_MITRAS_ONLINE)
+    pipe.del(VK_MITRAS_DEACTIVATED)
+
+    const now = new Date().toISOString()
+    if (onlineRows.length) {
+      pipe.sadd(VK_MITRAS_ONLINE, ...onlineRows.map((r) => r.mitra_id))
+      // Seed heartbeats with NOW so the first sweep after restart doesn't
+      // mass-offline. Mitras refresh on their next ping anyway.
+      for (const r of onlineRows) pipe.set(vkHeartbeatKey(r.mitra_id), now)
+      // Reset capacity for currently-online mitras; overlay real counts below.
+      // Offline mitras' stale capacity keys don't affect reads (SDIFF excludes them).
+      for (const r of onlineRows) pipe.set(vkCapacityKey(r.mitra_id), 0)
+    }
+    if (deactRows.length) {
+      pipe.sadd(VK_MITRAS_DEACTIVATED, ...deactRows.map((r) => r.id))
+    }
+    for (const r of capacityRows) pipe.set(vkCapacityKey(r.mitra_id), r.c)
+    await pipe.exec()
+
+    console.log(
+      `[valkey-mirror] seed: ${onlineRows.length} online, ${deactRows.length} deactivated, ${capacityRows.length} with active sessions`,
+    )
+  } catch (err) {
+    console.error('[valkey-mirror] seed failed:', err)
+  }
 }
 
-// Subscribe once at module load so other-instance config updates also bust this cache.
-// Single-instance: the local mutator already invalidates, so this is a no-op extra.
+// Re-seed on every Valkey reconnect (fires on initial connect too).
+onValkeyReady(seedFromPostgres)
+
+// --- Beacon snapshot cache (Valkey-backed, cluster-shared) ---
+// `availability:snapshot` JSON `{available, count}`, TTL 10s. All backend
+// instances share the same cache: one Valkey-driven compute per 10s
+// cluster-wide regardless of how many customers are polling.
+const AVAILABILITY_CACHE_KEY = 'availability:snapshot'
+const AVAILABILITY_TTL_SECONDS = 10
+
+export const invalidateAvailabilityCache = async () => {
+  try {
+    await valkey.del(AVAILABILITY_CACHE_KEY)
+  } catch (err) {
+    console.error('[valkey-mirror] invalidateAvailabilityCache failed:', err)
+  }
+}
+
+// Bust the shared cache when CC changes max_customers_per_mitra (any instance).
 let _subscribed = false
 const ensureSubscribed = () => {
   if (_subscribed) return
@@ -43,6 +99,39 @@ export const ensureStatusRow = async (mitraId) => {
   `
 }
 
+// Best-effort Valkey writer. Postgres remains source of truth; a Valkey hiccup
+// shouldn't fail the originating request — the reconciliation sweep heals drift.
+const tryValkey = async (fn, label) => {
+  try { await fn() } catch (err) {
+    console.error(`[valkey-mirror] ${label} failed:`, err)
+  }
+}
+
+// Recompute `mitra:capacity:<id>` from chat_sessions truth. Called after every
+// chat_session state change that could affect a mitra's occupied-slot count.
+// Recompute-from-truth avoids the bookkeeping risks of per-transition INCR/DECR
+// (double-counts, missed transitions across all the UPDATE sites in pairing,
+// closure, extension, session-timer, session services).
+export const recomputeCapacityForMitra = async (mitraId) => {
+  if (!mitraId) return
+  const [row] = await sql`
+    SELECT COUNT(*)::int AS c FROM chat_sessions
+    WHERE mitra_id = ${mitraId}
+      AND status IN (${SessionStatus.ACTIVE}, ${SessionStatus.PENDING_PAYMENT})
+  `
+  await tryValkey(
+    () => valkey.set(vkCapacityKey(mitraId), row.c),
+    `recomputeCapacity ${mitraId}`,
+  )
+}
+
+// Lookup mitra_id from the session, then recompute. Use this from UPDATE sites
+// where the session's mitra_id may not be in local scope.
+export const recomputeCapacityBySession = async (sessionId) => {
+  const [row] = await sql`SELECT mitra_id FROM chat_sessions WHERE id = ${sessionId}`
+  if (row?.mitra_id) await recomputeCapacityForMitra(row.mitra_id)
+}
+
 export const setOnline = async (mitraId) => {
   await ensureStatusRow(mitraId)
   const now = new Date()
@@ -54,6 +143,14 @@ export const setOnline = async (mitraId) => {
   await sql`
     INSERT INTO mitra_online_logs (mitra_id, status) VALUES (${mitraId}, 'online')
   `
+
+  await tryValkey(async () => {
+    const pipe = valkey.pipeline()
+    pipe.sadd(VK_MITRAS_ONLINE, mitraId)
+    pipe.set(vkHeartbeatKey(mitraId), now.toISOString())
+    await pipe.exec()
+  }, `setOnline ${mitraId}`)
+
   invalidateAvailabilityCache()
 }
 
@@ -73,16 +170,32 @@ export const setOffline = async (mitraId) => {
   await sql`
     INSERT INTO mitra_online_logs (mitra_id, status) VALUES (${mitraId}, 'offline')
   `
+
+  await tryValkey(async () => {
+    const pipe = valkey.pipeline()
+    pipe.srem(VK_MITRAS_ONLINE, mitraId)
+    pipe.del(vkHeartbeatKey(mitraId))
+    await pipe.exec()
+  }, `setOffline ${mitraId}`)
+
   invalidateAvailabilityCache()
 }
 
+// Heartbeat hot path: Valkey-only. Per-ping Postgres UPDATE eliminated; the
+// 60s batched heartbeat-mirror job (mirrorHeartbeatsToPostgres) writes
+// `last_heartbeat_at` to Postgres for forensics/restart safety.
+//
+// NOTE: there is intentionally no `is_online = true` gate here (the old SQL
+// UPDATE had one). The Valkey SET is global; if a mitra heartbeats while
+// `is_online=false` in Postgres, their TTL key gets refreshed but they're
+// still not in `mitras:online`, so blast eligibility is unchanged. The
+// reconciliation sweep will clean up the orphan heartbeat key.
 export const heartbeat = async (mitraId) => {
-  const now = new Date()
-  await sql`
-    UPDATE mitra_online_status
-    SET last_heartbeat_at = ${now}, updated_at = ${now}
-    WHERE mitra_id = ${mitraId} AND is_online = true
-  `
+  const now = new Date().toISOString()
+  await tryValkey(
+    () => valkey.set(vkHeartbeatKey(mitraId), now),
+    `heartbeat ${mitraId}`,
+  )
 }
 
 export const getStatus = async (mitraId) => {
@@ -130,39 +243,95 @@ export const getOnlineLogs = async (mitraId, { page = 1, limit = 50 } = {}) => {
   return { items, total: Number(count), page, limit }
 }
 
+// Valkey-driven: enumerate mitras:online, read each heartbeat timestamp from
+// Valkey, find stales, then bulk-flip Postgres + clean up Valkey.
+//
+// Failure semantics: if any Valkey op throws, the sweep aborts entirely. We
+// never mass-offline mitras via a Postgres scan because Valkey is unreachable
+// — that would risk false-offlining a fleet during a Valkey hiccup.
 export const autoOfflineStaleMitras = async () => {
   const pingConfig = await getMitraPingConfig()
-
-  // If ping is not required, skip the auto-offline sweep entirely
   if (!pingConfig.require_ping) return 0
 
-  // stale_after_seconds is the operator-facing knob — what they set is what
-  // they get. No multiplier, no implicit "tolerate N missed heartbeats"
-  // contract baked in. The CC PATCH validates that the value is >= the env-
-  // driven heartbeat cadence so single missed pings can't flip a mitra
-  // offline.
-  const staleSeconds = pingConfig.stale_after_seconds
-  const stale = await sql`
-    UPDATE mitra_online_status
-    SET is_online = false, last_offline_at = NOW(), updated_at = NOW()
-    WHERE is_online = true
-      AND last_heartbeat_at < NOW() - ${staleSeconds + ' seconds'}::interval
-    RETURNING mitra_id
-  `
-
-  for (const row of stale) {
-    await sql`
-      INSERT INTO mitra_online_logs (mitra_id, status) VALUES (${row.mitra_id}, 'offline')
-    `
+  let onlineIds, heartbeatValues
+  try {
+    onlineIds = await valkey.smembers(VK_MITRAS_ONLINE)
+    if (!onlineIds.length) return 0
+    const pipe = valkey.pipeline()
+    for (const id of onlineIds) pipe.get(vkHeartbeatKey(id))
+    const results = await pipe.exec()
+    heartbeatValues = results.map((r) => r[1])
+  } catch (err) {
+    console.warn('[auto-offline] valkey unavailable, skipping this tick:', err.message)
+    return 0
   }
 
-  // Capacity may have changed (mitra went offline) — invalidate the customer-facing
-  // availability cache so the next poll reflects reality.
-  if (stale.length > 0) invalidateAvailabilityCache()
+  const cutoff = Date.now() - pingConfig.stale_after_seconds * 1000
+  const stale = []
+  for (let i = 0; i < onlineIds.length; i++) {
+    const ts = heartbeatValues[i]
+    if (!ts || Date.parse(ts) < cutoff) stale.push(onlineIds[i])
+  }
+  if (!stale.length) return 0
 
+  await sql`
+    UPDATE mitra_online_status
+    SET is_online = false, last_offline_at = NOW(), updated_at = NOW()
+    WHERE mitra_id = ANY(${sql.array(stale)}::uuid[]) AND is_online = true
+  `
+  for (const id of stale) {
+    await sql`INSERT INTO mitra_online_logs (mitra_id, status) VALUES (${id}, 'offline')`
+  }
+  await tryValkey(async () => {
+    const cleanup = valkey.pipeline()
+    cleanup.srem(VK_MITRAS_ONLINE, ...stale)
+    for (const id of stale) cleanup.del(vkHeartbeatKey(id))
+    await cleanup.exec()
+  }, `auto-offline cleanup (${stale.length} stale)`)
+
+  invalidateAvailabilityCache()
   return stale.length
 }
 
+// Batched mirror: Valkey heartbeat timestamps → Postgres `last_heartbeat_at`.
+// Runs every HEARTBEAT_MIRROR_INTERVAL_SECONDS (default 60). One UNNEST UPDATE
+// regardless of online count. Idempotent — latest timestamp wins; multiple
+// instances running concurrently is fine (no leader election needed).
+export const mirrorHeartbeatsToPostgres = async () => {
+  let onlineIds, heartbeatValues
+  try {
+    onlineIds = await valkey.smembers(VK_MITRAS_ONLINE)
+    if (!onlineIds.length) return 0
+    const pipe = valkey.pipeline()
+    for (const id of onlineIds) pipe.get(vkHeartbeatKey(id))
+    const results = await pipe.exec()
+    heartbeatValues = results.map((r) => r[1])
+  } catch (err) {
+    console.warn('[heartbeat-mirror] valkey unavailable, skipping:', err.message)
+    return 0
+  }
+
+  const ids = []
+  const ts = []
+  for (let i = 0; i < onlineIds.length; i++) {
+    if (heartbeatValues[i]) {
+      ids.push(onlineIds[i])
+      ts.push(heartbeatValues[i])
+    }
+  }
+  if (!ids.length) return 0
+
+  await sql`
+    UPDATE mitra_online_status m
+    SET last_heartbeat_at = u.ts::timestamptz, updated_at = NOW()
+    FROM (
+      SELECT * FROM UNNEST(${sql.array(ids)}::uuid[], ${sql.array(ts)}::text[]) AS t(mitra_id, ts)
+    ) u
+    WHERE m.mitra_id = u.mitra_id
+  `
+  return ids.length
+}
+
 /**
  * Customer-home availability check, cached in-memory for AVAILABILITY_TTL_MS.
  *
@@ -178,12 +347,33 @@ export const autoOfflineStaleMitras = async () => {
  * sets/hashes (matching the existing memory item "Session Timer Scaling"); the contract
  * of this function — Valkey/cache reads only on the hot path — stays the same.
  */
-export const countAvailableMitrasFromCache = async () => {
-  const now = Date.now()
-  if (availabilityCache && availabilityCache.expiresAt > now) {
-    return { available: availabilityCache.available, count: availabilityCache.count }
-  }
+const computeAvailabilityFromValkey = async () => {
+  const { max_customers_per_mitra } = await getMaxCustomersPerMitra()
+  const { stale_after_seconds } = await getMitraPingConfig()
 
+  const candidates = await valkey.sdiff(VK_MITRAS_ONLINE, VK_MITRAS_DEACTIVATED)
+  if (!candidates.length) return { available: false, count: 0 }
+
+  const pipe = valkey.pipeline()
+  for (const id of candidates) {
+    pipe.get(vkCapacityKey(id))
+    pipe.get(vkHeartbeatKey(id))
+  }
+  const results = await pipe.exec()
+
+  const cutoff = Date.now() - stale_after_seconds * 1000
+  let count = 0
+  for (let i = 0; i < candidates.length; i++) {
+    const capacity = Number(results[i * 2][1] ?? 0)
+    const heartbeat = results[i * 2 + 1][1]
+    if (capacity >= max_customers_per_mitra) continue
+    if (!heartbeat || Date.parse(heartbeat) < cutoff) continue
+    count++
+  }
+  return { available: count > 0, count }
+}
+
+const computeAvailabilityFromPostgres = async () => {
   const { max_customers_per_mitra } = await getMaxCustomersPerMitra()
   const [{ count }] = await sql`
     SELECT COUNT(*)::int AS count
@@ -197,26 +387,42 @@ export const countAvailableMitrasFromCache = async () => {
           AND cs.status IN (${SessionStatus.ACTIVE}, ${SessionStatus.PENDING_PAYMENT})
       ) < ${max_customers_per_mitra}
   `
+  return { available: count > 0, count }
+}
 
-  const available = count > 0
-  availabilityCache = {
-    available,
-    count,
-    expiresAt: now + AVAILABILITY_TTL_MS,
+export const countAvailableMitrasFromCache = async () => {
+  try {
+    const cached = await valkey.get(AVAILABILITY_CACHE_KEY)
+    if (cached) return JSON.parse(cached)
+
+    const snapshot = await computeAvailabilityFromValkey()
+    await valkey.getValkeyClient().setex(AVAILABILITY_CACHE_KEY, AVAILABILITY_TTL_SECONDS, JSON.stringify(snapshot))
+    return snapshot
+  } catch (err) {
+    console.warn('[countAvailableMitras] valkey unavailable, falling back to Postgres:', err.message)
+    return computeAvailabilityFromPostgres()
   }
-  return { available, count }
 }
 
 /**
- * Mitra-online check for use during pairing/extension safeguards.
- * Combines the Valkey-mirrored online flag (Postgres mitra_online_status today) with
- * the WebSocket-connected check. Never use "in-session" as a proxy for "online".
+ * Mitra-reachable check: in `mitras:online` SET AND heartbeat is fresh.
+ * Falls back to a Postgres `is_online` read if Valkey is unreachable; the
+ * fallback skips the heartbeat-freshness check (sweep takes care of stale rows
+ * within `stale_after_seconds + sweep_cadence`).
  */
 export const isMitraReachable = async (mitraId) => {
-  const [row] = await sql`
-    SELECT is_online FROM mitra_online_status WHERE mitra_id = ${mitraId}
-  `
-  return Boolean(row?.is_online)
+  try {
+    const inSet = await valkey.sismember(VK_MITRAS_ONLINE, mitraId)
+    if (!inSet) return false
+    const heartbeat = await valkey.get(vkHeartbeatKey(mitraId))
+    if (!heartbeat) return false
+    const { stale_after_seconds } = await getMitraPingConfig()
+    return Date.parse(heartbeat) >= Date.now() - stale_after_seconds * 1000
+  } catch (err) {
+    console.warn('[isMitraReachable] valkey unavailable, falling back to DB:', err.message)
+    const [row] = await sql`SELECT is_online FROM mitra_online_status WHERE mitra_id = ${mitraId}`
+    return Boolean(row?.is_online)
+  }
 }
 
 /**
diff --git a/backend/src/services/mitra.service.js b/backend/src/services/mitra.service.js
index 2daf15c..089675c 100644
--- a/backend/src/services/mitra.service.js
+++ b/backend/src/services/mitra.service.js
@@ -1,4 +1,8 @@
 import { getDb } from '../db/client.js'
+import * as valkey from '../plugins/valkey.js'
+import { VK_MITRAS_DEACTIVATED } from './mitra-status.service.js'
+import { revokeAllSessionsForUser } from './token.service.js'
+import { UserType } from '../constants.js'
 
 const sql = getDb()
 
@@ -36,6 +40,24 @@ export const updateMitraStatus = async (id, is_active) => {
     RETURNING id, is_active
   `
   if (!mitra) throw Object.assign(new Error('Mitra not found'), { code: 'NOT_FOUND', statusCode: 404 })
+
+  // Deactivation also revokes all auth_sessions so the next token refresh fails
+  // (bounds the "ghost online" window to access-token TTL across all routes,
+  // not just heartbeat). See requirement/valkey-online-mirror-plan.md.
+  if (!is_active) {
+    await revokeAllSessionsForUser({ userType: UserType.MITRA, userId: id })
+  }
+
+  try {
+    if (is_active) {
+      await valkey.srem(VK_MITRAS_DEACTIVATED, id)
+    } else {
+      await valkey.sadd(VK_MITRAS_DEACTIVATED, id)
+    }
+  } catch (err) {
+    console.error(`[valkey-mirror] updateMitraStatus ${id} failed:`, err)
+  }
+
   return mitra
 }
 
diff --git a/backend/src/services/pairing.service.js b/backend/src/services/pairing.service.js
index 1816ca1..d4e1215 100644
--- a/backend/src/services/pairing.service.js
+++ b/backend/src/services/pairing.service.js
@@ -1,11 +1,13 @@
 import { getDb } from '../db/client.js'
-import { getMaxCustomersPerMitra, getPairingBlastTimeoutSeconds, getReturningChatConfirmationTimeoutSeconds } from './config.service.js'
+import * as valkey from '../plugins/valkey.js'
+import { VK_MITRAS_ONLINE, VK_MITRAS_DEACTIVATED, vkCapacityKey, vkHeartbeatKey } from './mitra-status.service.js'
+import { getMaxCustomersPerMitra, getPairingBlastTimeoutSeconds, getReturningChatConfirmationTimeoutSeconds, getMitraPingConfig } from './config.service.js'
 import { sendToUser } from '../plugins/websocket.js'
 import { sendPushNotification } from './notification.service.js'
 import { startSessionTimer } from './session-timer.service.js'
 import { startSessionListener } from './chat-handler.service.js'
 import { consumePaymentSession, failPaymentSession, getPaymentSession, recordIntermediateFailure } from './payment.service.js'
-import { isMitraReachable, isMitraInActiveSessionWithCustomer, getMitraActiveSessionCount } from './mitra-status.service.js'
+import { isMitraReachable, isMitraInActiveSessionWithCustomer, getMitraActiveSessionCount, recomputeCapacityForMitra, recomputeCapacityBySession } from './mitra-status.service.js'
 import {
   UserType,
   SessionStatus,
@@ -76,10 +78,37 @@ const notifyCustomer = async (customerId, data) => {
   }
 }
 
-export const findAvailableMitras = async () => {
+// Valkey-driven: SDIFF(online, deactivated) → for each candidate, pipelined
+// GET capacity + heartbeat, then filter by capacity gate + heartbeat freshness.
+// Postgres fallback runs if any Valkey op throws (full JOIN as before).
+const findAvailableMitrasFromValkey = async () => {
+  const { max_customers_per_mitra } = await getMaxCustomersPerMitra()
+  const { stale_after_seconds } = await getMitraPingConfig()
+
+  const candidates = await valkey.sdiff(VK_MITRAS_ONLINE, VK_MITRAS_DEACTIVATED)
+  if (!candidates.length) return []
+
+  const pipe = valkey.pipeline()
+  for (const id of candidates) {
+    pipe.get(vkCapacityKey(id))
+    pipe.get(vkHeartbeatKey(id))
+  }
+  const results = await pipe.exec()
+
+  const cutoff = Date.now() - stale_after_seconds * 1000
+  const eligible = []
+  for (let i = 0; i < candidates.length; i++) {
+    const capacity = Number(results[i * 2][1] ?? 0)
+    const heartbeat = results[i * 2 + 1][1]
+    if (capacity >= max_customers_per_mitra) continue
+    if (!heartbeat || Date.parse(heartbeat) < cutoff) continue
+    eligible.push({ id: candidates[i], active_session_count: capacity })
+  }
+  return eligible
+}
+
+const findAvailableMitrasFromPostgres = async () => {
   const { max_customers_per_mitra } = await getMaxCustomersPerMitra()
-  // Project active_session_count alongside the mitra row so the blast loop doesn't
-  // need a per-mitra COUNT roundtrip later.
   const mitras = await sql`
     SELECT m.id, m.display_name, sub.active_session_count
     FROM mitras m
@@ -96,6 +125,15 @@ export const findAvailableMitras = async () => {
   return mitras
 }
 
+export const findAvailableMitras = async () => {
+  try {
+    return await findAvailableMitrasFromValkey()
+  } catch (err) {
+    console.warn('[findAvailableMitras] valkey unavailable, falling back to Postgres:', err.message)
+    return findAvailableMitrasFromPostgres()
+  }
+}
+
 /**
  * Validate that a payment session is owned by the customer, confirmed, and not yet consumed.
  * Throws on mismatch. Returns the loaded payment session row.
@@ -414,6 +452,10 @@ export const acceptPairingRequest = async (sessionId, mitraId) => {
     })
   }
 
+  // Mitra now occupies a capacity slot (PENDING_PAYMENT counts per
+  // findAvailableMitras predicate). Mirror to Valkey.
+  await recomputeCapacityForMitra(mitraId)
+
   // Mark this mitra's notification as accepted
   await sql`
     UPDATE chat_request_notifications
diff --git a/backend/src/services/session-timer.service.js b/backend/src/services/session-timer.service.js
index 1d45581..d2153cb 100644
--- a/backend/src/services/session-timer.service.js
+++ b/backend/src/services/session-timer.service.js
@@ -2,6 +2,7 @@ import { getDb } from '../db/client.js'
 import { publish } from '../plugins/valkey.js'
 import { sendToSessionParticipant } from '../plugins/websocket.js'
 import { sendPushNotification } from './notification.service.js'
+import { recomputeCapacityForMitra } from './mitra-status.service.js'
 import { UserType, SessionStatus, WsMessage, EndedBy } from '../constants.js'
 
 const sql = getDb()
@@ -152,6 +153,7 @@ const onSessionExpired = async (sessionId) => {
     RETURNING id, customer_id, mitra_id
   `
   if (!session) return
+  await recomputeCapacityForMitra(session.mitra_id)
 
   // Notify customer — sees extend/close dialog; FCM fallback if WebSocket is down
   const expiredData = { type: WsMessage.SESSION_EXPIRED, session_id: sessionId }
@@ -207,9 +209,10 @@ const autoCompleteIfStillClosing = async (sessionId) => {
           ended_at = COALESCE(ended_at, NOW()),
           ended_by = ${EndedBy.SYSTEM_AUTO_CLOSE}
       WHERE id = ${sessionId} AND status = ${SessionStatus.CLOSING}
-      RETURNING id
+      RETURNING id, mitra_id
     `
     if (!updated) return
+    await recomputeCapacityForMitra(updated.mitra_id)
 
     const data = { type: WsMessage.SESSION_COMPLETED, session_id: sessionId }
     sendToSessionParticipant(sessionId, UserType.CUSTOMER, data)
diff --git a/backend/src/services/session.service.js b/backend/src/services/session.service.js
index 5514463..81b4dfc 100644
--- a/backend/src/services/session.service.js
+++ b/backend/src/services/session.service.js
@@ -1,5 +1,6 @@
 import { getDb } from '../db/client.js'
 import { publish } from '../plugins/valkey.js'
+import { recomputeCapacityForMitra } from './mitra-status.service.js'
 import { UserType, SessionStatus, MessageStatus, WsMessage } from '../constants.js'
 
 const sql = getDb()
@@ -48,6 +49,7 @@ export const endSession = async (sessionId, endedBy, userId) => {
       code: 'SESSION_NOT_ACTIVE', statusCode: 409,
     })
   }
+  await recomputeCapacityForMitra(session.mitra_id)
 
   // Notify both parties
   await publish(`session:${sessionId}:status`, {
@@ -91,6 +93,11 @@ export const rerouteSession = async (sessionId, newMitraId) => {
     WHERE id = ${sessionId}
     RETURNING id, customer_id, mitra_id, status
   `
+  // Both mitras' capacity flipped — recompute both.
+  await Promise.all([
+    recomputeCapacityForMitra(oldMitraId),
+    recomputeCapacityForMitra(newMitraId),
+  ])
 
   const [newMitra] = await sql`
     SELECT display_name FROM mitras WHERE id = ${newMitraId}
diff --git a/backend/test/helpers/db.js b/backend/test/helpers/db.js
index 41e31fa..39717d5 100644
--- a/backend/test/helpers/db.js
+++ b/backend/test/helpers/db.js
@@ -1,4 +1,5 @@
 import { getDb } from '../../src/db/client.js'
+import { flushTestDb } from './valkey.js'
 
 /**
  * Single shared sql client used by tests. Same singleton the services use, since
@@ -37,6 +38,9 @@ export const resetDb = async () => {
   const sql = db()
   // RESTART IDENTITY is a no-op for UUID PKs but cheap; CASCADE handles any future FK additions.
   await sql.unsafe(`TRUNCATE TABLE ${TRUNCATE_TABLES.join(', ')} RESTART IDENTITY CASCADE`)
+  // Flush Valkey availability state so each test starts hermetic. Fixtures
+  // (createMitra etc.) re-seed Valkey alongside their Postgres writes.
+  await flushTestDb()
 }
 
 /**
diff --git a/backend/test/helpers/fixtures.js b/backend/test/helpers/fixtures.js
index 2388935..a200901 100644
--- a/backend/test/helpers/fixtures.js
+++ b/backend/test/helpers/fixtures.js
@@ -1,5 +1,6 @@
 import { randomUUID } from 'node:crypto'
 import { db, resetAppConfig } from './db.js'
+import { getTestValkey } from './valkey.js'
 
 /**
  * Insert a customer row. Defaults to the schema after the Phase 3.4 auth rewrite
@@ -47,6 +48,19 @@ export const createMitra = async ({
       ON CONFLICT (mitra_id) DO UPDATE
         SET is_online = true, last_online_at = ${now}, last_heartbeat_at = ${now}, updated_at = ${now}
     `
+    // Mirror to Valkey so findAvailableMitras (Valkey-driven) sees this mitra.
+    // resetDb already FLUSHDBs Valkey, so seeding here per-mitra keeps tests
+    // hermetic without depending on production's startup seed.
+    const v = getTestValkey()
+    await v.multi()
+      .sadd('mitras:online', id)
+      .set(`mitra:heartbeat:${id}`, now.toISOString())
+      .set(`mitra:capacity:${id}`, 0)
+      .exec()
+  }
+  if (!isActive) {
+    const v = getTestValkey()
+    await v.sadd('mitras:deactivated', id)
   }
   return row
 }
diff --git a/backend/test/services/mitra-status.valkey-mirror.test.js b/backend/test/services/mitra-status.valkey-mirror.test.js
new file mode 100644
index 0000000..e99b1e3
--- /dev/null
+++ b/backend/test/services/mitra-status.valkey-mirror.test.js
@@ -0,0 +1,494 @@
+import { describe, it, expect, beforeAll, beforeEach, afterAll, vi } from 'vitest'
+
+/**
+ * Integration tests for the Valkey availability mirror
+ * (requirement/valkey-online-mirror-plan.md).
+ *
+ * Real Postgres + real Valkey via test/setup.js — no mocks. We assert on both
+ * stores' state after each operation to catch missed mirrors or order bugs.
+ */
+
+vi.mock('../../src/plugins/websocket.js', () => ({
+  sendToUser: vi.fn(() => false),
+  sendToSessionParticipant: vi.fn(() => false),
+  registerWebSocketPlugin: vi.fn(),
+  registerWebSocketRoute: vi.fn(),
+  isUserOnlineWs: vi.fn(() => false),
+  getSessionConnections: vi.fn(() => ({})),
+}))
+
+vi.mock('../../src/services/notification.service.js', () => ({
+  sendPushNotification: vi.fn(async () => true),
+  registerDeviceToken: vi.fn(async () => {}),
+}))
+
+const valkey = await import('../../src/plugins/valkey.js')
+const {
+  setOnline,
+  setOffline,
+  heartbeat,
+  isMitraReachable,
+  recomputeCapacityForMitra,
+  recomputeCapacityBySession,
+  seedFromPostgres,
+  autoOfflineStaleMitras,
+  mirrorHeartbeatsToPostgres,
+  countAvailableMitrasFromCache,
+  invalidateAvailabilityCache,
+  VK_MITRAS_ONLINE,
+  VK_MITRAS_DEACTIVATED,
+  vkCapacityKey,
+  vkHeartbeatKey,
+} = await import('../../src/services/mitra-status.service.js')
+const { updateMitraStatus } = await import('../../src/services/mitra.service.js')
+const { findAvailableMitras, acceptPairingRequest, createPairingRequest } = await import('../../src/services/pairing.service.js')
+const { createPaymentSession, confirmPaymentSession } = await import('../../src/services/payment.service.js')
+const { db, resetDb, resetAppConfig } = await import('../helpers/db.js')
+const { getTestValkey } = await import('../helpers/valkey.js')
+const { createCustomer, createMitra } = await import('../helpers/fixtures.js')
+const { SessionStatus, UserType } = await import('../../src/constants.js')
+
+const v = () => getTestValkey()
+
+describe('mitra-status valkey mirror', () => {
+  beforeAll(async () => {
+    await resetAppConfig()
+  })
+
+  beforeEach(async () => {
+    await resetDb()
+  })
+
+  // ---------- Seed ----------
+
+  describe('seedFromPostgres', () => {
+    it('populates mitras:online from is_online=true rows', async () => {
+      const m1 = await createMitra({ callName: 'M1', isOnline: true })
+      const m2 = await createMitra({ callName: 'M2', isOnline: true })
+      await createMitra({ callName: 'M3', isOnline: false })
+
+      await v().flushdb()
+      await seedFromPostgres()
+
+      const members = await v().smembers(VK_MITRAS_ONLINE)
+      expect(members.sort()).toEqual([m1.id, m2.id].sort())
+    })
+
+    it('seeds mitras:deactivated from is_active=false', async () => {
+      const m = await createMitra({ callName: 'Dead', isActive: false })
+      await createMitra({ callName: 'Alive', isActive: true })
+
+      await v().flushdb()
+      await seedFromPostgres()
+
+      const members = await v().smembers(VK_MITRAS_DEACTIVATED)
+      expect(members).toEqual([m.id])
+    })
+
+    it('seeds heartbeat keys for online mitras with current timestamp', async () => {
+      const m = await createMitra({ callName: 'Live', isOnline: true })
+      await v().flushdb()
+
+      const before = Date.now()
+      await seedFromPostgres()
+      const after = Date.now()
+
+      const ts = await v().get(vkHeartbeatKey(m.id))
+      expect(ts).toBeTruthy()
+      const seeded = Date.parse(ts)
+      expect(seeded).toBeGreaterThanOrEqual(before)
+      expect(seeded).toBeLessThanOrEqual(after)
+    })
+
+    it('seeds capacity counters from chat_sessions', async () => {
+      const c = await createCustomer({ callName: 'C' })
+      const m = await createMitra({ callName: 'M', isOnline: true })
+      const sql = db()
+      await sql`
+        INSERT INTO chat_sessions (customer_id, mitra_id, status)
+        VALUES (${c.id}, ${m.id}, ${SessionStatus.ACTIVE})
+      `
+
+      await v().flushdb()
+      await seedFromPostgres()
+
+      expect(await v().get(vkCapacityKey(m.id))).toBe('1')
+    })
+
+    it('is idempotent — running twice yields the same state', async () => {
+      const m = await createMitra({ callName: 'Idem', isOnline: true })
+      await seedFromPostgres()
+      const first = {
+        online: (await v().smembers(VK_MITRAS_ONLINE)).sort(),
+        heartbeat: await v().get(vkHeartbeatKey(m.id)),
+      }
+      await seedFromPostgres()
+      const second = {
+        online: (await v().smembers(VK_MITRAS_ONLINE)).sort(),
+        heartbeat: await v().get(vkHeartbeatKey(m.id)),
+      }
+      expect(second.online).toEqual(first.online)
+      // Heartbeat is reseeded with NOW each call — must be >= first
+      expect(Date.parse(second.heartbeat)).toBeGreaterThanOrEqual(Date.parse(first.heartbeat))
+    })
+  })
+
+  // ---------- setOnline / setOffline ----------
+
+  describe('setOnline / setOffline write-through', () => {
+    it('setOnline adds to mitras:online + writes heartbeat key', async () => {
+      const m = await createMitra({ callName: 'Toggle', isOnline: false })
+      await v().flushdb()
+
+      await setOnline(m.id)
+
+      expect(await v().sismember(VK_MITRAS_ONLINE, m.id)).toBe(1)
+      expect(await v().get(vkHeartbeatKey(m.id))).toBeTruthy()
+    })
+
+    it('setOffline removes from mitras:online + deletes heartbeat key', async () => {
+      const m = await createMitra({ callName: 'Toggle', isOnline: true })
+      await setOnline(m.id) // ensure heartbeat key exists
+
+      await setOffline(m.id)
+
+      expect(await v().sismember(VK_MITRAS_ONLINE, m.id)).toBe(0)
+      expect(await v().get(vkHeartbeatKey(m.id))).toBeNull()
+    })
+
+    it('setOffline is no-op when mitra was already offline', async () => {
+      const m = await createMitra({ callName: 'OffAlready', isOnline: false })
+      const sql = db()
+      const beforeLogs = await sql`SELECT COUNT(*)::int AS c FROM mitra_online_logs WHERE mitra_id=${m.id}`
+
+      await setOffline(m.id)
+
+      const afterLogs = await sql`SELECT COUNT(*)::int AS c FROM mitra_online_logs WHERE mitra_id=${m.id}`
+      expect(afterLogs[0].c).toBe(beforeLogs[0].c)
+    })
+  })
+
+  // ---------- heartbeat ----------
+
+  describe('heartbeat (Valkey-only)', () => {
+    it('writes Valkey timestamp without touching Postgres last_heartbeat_at', async () => {
+      const m = await createMitra({ callName: 'Pinger', isOnline: true })
+      const sql = db()
+      const [before] = await sql`SELECT last_heartbeat_at FROM mitra_online_status WHERE mitra_id=${m.id}`
+      const pgBefore = before.last_heartbeat_at
+
+      // Make sure subsequent NOW() would differ
+      await new Promise(r => setTimeout(r, 50))
+      await heartbeat(m.id)
+
+      const [after] = await sql`SELECT last_heartbeat_at FROM mitra_online_status WHERE mitra_id=${m.id}`
+      // Postgres untouched
+      expect(after.last_heartbeat_at).toEqual(pgBefore)
+      // Valkey updated
+      const ts = await v().get(vkHeartbeatKey(m.id))
+      expect(ts).toBeTruthy()
+      expect(Date.parse(ts)).toBeGreaterThan(pgBefore.getTime())
+    })
+
+    it('advances the heartbeat timestamp on each call', async () => {
+      const m = await createMitra({ callName: 'P', isOnline: true })
+      await heartbeat(m.id)
+      const t1 = await v().get(vkHeartbeatKey(m.id))
+
+      await new Promise(r => setTimeout(r, 20))
+      await heartbeat(m.id)
+      const t2 = await v().get(vkHeartbeatKey(m.id))
+
+      expect(Date.parse(t2)).toBeGreaterThan(Date.parse(t1))
+    })
+  })
+
+  // ---------- isMitraReachable ----------
+
+  describe('isMitraReachable', () => {
+    it('returns true for online mitra with fresh heartbeat', async () => {
+      const m = await createMitra({ callName: 'Reach', isOnline: true })
+      expect(await isMitraReachable(m.id)).toBe(true)
+    })
+
+    it('returns false when mitra is not in mitras:online', async () => {
+      const m = await createMitra({ callName: 'NoReach', isOnline: false })
+      expect(await isMitraReachable(m.id)).toBe(false)
+    })
+
+    it('returns false when heartbeat is stale', async () => {
+      const m = await createMitra({ callName: 'Stale', isOnline: true })
+      // Force stale heartbeat (one hour ago)
+      const ancient = new Date(Date.now() - 3_600_000).toISOString()
+      await v().set(vkHeartbeatKey(m.id), ancient)
+
+      expect(await isMitraReachable(m.id)).toBe(false)
+    })
+  })
+
+  // ---------- recomputeCapacity ----------
+
+  describe('recomputeCapacityForMitra', () => {
+    it('counts ACTIVE + PENDING_PAYMENT sessions', async () => {
+      const c = await createCustomer({ callName: 'C' })
+      const c2 = await createCustomer({ callName: 'C2' })
+      const m = await createMitra({ callName: 'Cap', isOnline: true })
+      const sql = db()
+      await sql`
+        INSERT INTO chat_sessions (customer_id, mitra_id, status)
+        VALUES (${c.id}, ${m.id}, ${SessionStatus.ACTIVE}),
+               (${c2.id}, ${m.id}, ${SessionStatus.PENDING_PAYMENT})
+      `
+      await recomputeCapacityForMitra(m.id)
+      expect(await v().get(vkCapacityKey(m.id))).toBe('2')
+    })
+
+    it('excludes ended/closing/extending sessions', async () => {
+      const c = await createCustomer({ callName: 'C' })
+      const m = await createMitra({ callName: 'Cap', isOnline: true })
+      const sql = db()
+      await sql`
+        INSERT INTO chat_sessions (customer_id, mitra_id, status)
+        VALUES (${c.id}, ${m.id}, ${SessionStatus.COMPLETED})
+      `
+      await recomputeCapacityForMitra(m.id)
+      expect(await v().get(vkCapacityKey(m.id))).toBe('0')
+    })
+
+    it('no-op when mitraId is null/undefined', async () => {
+      await recomputeCapacityForMitra(null)  // should not throw
+      await recomputeCapacityForMitra(undefined)
+    })
+  })
+
+  // ---------- findAvailableMitras ----------
+
+  describe('findAvailableMitras (Valkey-driven)', () => {
+    it('returns online + not-deactivated + under-capacity + fresh-heartbeat mitras', async () => {
+      const ok = await createMitra({ callName: 'OK', isOnline: true })
+      const deact = await createMitra({ callName: 'Deact', isOnline: true, isActive: false })
+      const offline = await createMitra({ callName: 'Off', isOnline: false })
+      const stale = await createMitra({ callName: 'Stale', isOnline: true })
+      await v().set(vkHeartbeatKey(stale.id), new Date(Date.now() - 3_600_000).toISOString())
+
+      const result = await findAvailableMitras()
+      const ids = result.map(r => r.id).sort()
+      expect(ids).toEqual([ok.id].sort())
+      expect(result.find(r => r.id === ok.id).active_session_count).toBe(0)
+    })
+
+    it('excludes a mitra whose capacity is at max', async () => {
+      const m = await createMitra({ callName: 'AtCap', isOnline: true })
+      // max_customers_per_mitra default is 3
+      await v().set(vkCapacityKey(m.id), 3)
+
+      const result = await findAvailableMitras()
+      expect(result.find(r => r.id === m.id)).toBeUndefined()
+    })
+
+    it('returns capacity in the result for the blast loop', async () => {
+      const m = await createMitra({ callName: 'WithCap', isOnline: true })
+      await v().set(vkCapacityKey(m.id), 2)
+      const result = await findAvailableMitras()
+      expect(result.find(r => r.id === m.id).active_session_count).toBe(2)
+    })
+  })
+
+  // ---------- countAvailableMitrasFromCache ----------
+
+  describe('countAvailableMitrasFromCache (beacon)', () => {
+    it('caches the snapshot in Valkey with TTL', async () => {
+      await createMitra({ callName: 'On', isOnline: true })
+      await v().del('availability:snapshot')
+
+      const first = await countAvailableMitrasFromCache()
+      expect(first.available).toBe(true)
+      expect(first.count).toBe(1)
+
+      const cached = await v().get('availability:snapshot')
+      expect(cached).toBeTruthy()
+      expect(JSON.parse(cached)).toEqual(first)
+
+      const ttl = await v().ttl('availability:snapshot')
+      expect(ttl).toBeGreaterThan(0)
+      expect(ttl).toBeLessThanOrEqual(10)
+    })
+
+    it('returns cached snapshot on subsequent calls without recompute', async () => {
+      await createMitra({ callName: 'On', isOnline: true })
+      await countAvailableMitrasFromCache() // primes cache
+
+      // Manually corrupt SET to prove subsequent call reads cache, not Valkey state
+      await v().flushdb()
+      await v().set('availability:snapshot', JSON.stringify({ available: true, count: 42 }), 'EX', 10)
+
+      const result = await countAvailableMitrasFromCache()
+      expect(result.count).toBe(42)
+    })
+
+    it('invalidateAvailabilityCache deletes the snapshot', async () => {
+      await v().set('availability:snapshot', JSON.stringify({ available: true, count: 1 }), 'EX', 10)
+      await invalidateAvailabilityCache()
+      expect(await v().get('availability:snapshot')).toBeNull()
+    })
+  })
+
+  // ---------- autoOfflineStaleMitras ----------
+
+  describe('autoOfflineStaleMitras', () => {
+    it('flips Postgres + cleans Valkey for mitras with stale heartbeat', async () => {
+      const m = await createMitra({ callName: 'WillStale', isOnline: true })
+      const sql = db()
+
+      // Force stale heartbeat
+      await v().set(vkHeartbeatKey(m.id), new Date(Date.now() - 3_600_000).toISOString())
+
+      const count = await autoOfflineStaleMitras()
+      expect(count).toBe(1)
+
+      const [row] = await sql`SELECT is_online FROM mitra_online_status WHERE mitra_id=${m.id}`
+      expect(row.is_online).toBe(false)
+      expect(await v().sismember(VK_MITRAS_ONLINE, m.id)).toBe(0)
+      expect(await v().get(vkHeartbeatKey(m.id))).toBeNull()
+
+      const [log] = await sql`
+        SELECT status FROM mitra_online_logs
+        WHERE mitra_id=${m.id} ORDER BY timestamp DESC LIMIT 1
+      `
+      expect(log.status).toBe('offline')
+    })
+
+    it('no-op when no mitras are stale', async () => {
+      await createMitra({ callName: 'Fresh', isOnline: true })
+      const count = await autoOfflineStaleMitras()
+      expect(count).toBe(0)
+    })
+
+    it('no-op when require_ping=false', async () => {
+      const sql = db()
+      await sql`
+        UPDATE app_config SET value=${sql.json({ value: false })}
+        WHERE key='require_mitra_ping'
+      `
+      const m = await createMitra({ callName: 'WouldBeStale', isOnline: true })
+      await v().set(vkHeartbeatKey(m.id), new Date(Date.now() - 3_600_000).toISOString())
+
+      const count = await autoOfflineStaleMitras()
+      expect(count).toBe(0)
+      // Restore for other tests
+      await sql`
+        UPDATE app_config SET value=${sql.json({ value: true })}
+        WHERE key='require_mitra_ping'
+      `
+    })
+  })
+
+  // ---------- mirrorHeartbeatsToPostgres ----------
+
+  describe('mirrorHeartbeatsToPostgres', () => {
+    it('writes Valkey heartbeat timestamps to Postgres last_heartbeat_at in one batch', async () => {
+      const m1 = await createMitra({ callName: 'P1', isOnline: true })
+      const m2 = await createMitra({ callName: 'P2', isOnline: true })
+      const sql = db()
+
+      const ts = new Date(Date.now() - 2_000).toISOString()
+      await v().set(vkHeartbeatKey(m1.id), ts)
+      await v().set(vkHeartbeatKey(m2.id), ts)
+
+      const count = await mirrorHeartbeatsToPostgres()
+      expect(count).toBe(2)
+
+      const rows = await sql`
+        SELECT mitra_id, last_heartbeat_at FROM mitra_online_status
+        WHERE mitra_id IN (${m1.id}, ${m2.id})
+      `
+      for (const row of rows) {
+        expect(row.last_heartbeat_at.toISOString()).toBe(ts)
+      }
+    })
+
+    it('no-op when no mitras are online', async () => {
+      await v().del(VK_MITRAS_ONLINE)
+      const count = await mirrorHeartbeatsToPostgres()
+      expect(count).toBe(0)
+    })
+  })
+
+  // ---------- updateMitraStatus / revokeAllSessions ----------
+
+  describe('updateMitraStatus + auth_session revocation', () => {
+    it('deactivation adds to mitras:deactivated AND revokes all auth_sessions', async () => {
+      const m = await createMitra({ callName: 'Banned', isActive: true })
+      const sql = db()
+      const tokenHash = '$2b$10$abcdefghijklmnopqrstuv'
+      await sql`
+        INSERT INTO auth_sessions (user_type, user_id, refresh_token_hash, expires_at)
+        VALUES (${UserType.MITRA}, ${m.id}, ${tokenHash}, NOW() + INTERVAL '30 days')
+      `
+
+      await updateMitraStatus(m.id, false)
+
+      expect(await v().sismember(VK_MITRAS_DEACTIVATED, m.id)).toBe(1)
+      const [auth] = await sql`SELECT revoked_at FROM auth_sessions WHERE user_id=${m.id}`
+      expect(auth.revoked_at).not.toBeNull()
+    })
+
+    it('reactivation removes from mitras:deactivated', async () => {
+      const m = await createMitra({ callName: 'Pardoned', isActive: false })
+      await v().sadd(VK_MITRAS_DEACTIVATED, m.id)
+
+      await updateMitraStatus(m.id, true)
+
+      expect(await v().sismember(VK_MITRAS_DEACTIVATED, m.id)).toBe(0)
+    })
+  })
+
+  // ---------- E2E: blast lifecycle ----------
+
+  describe('end-to-end: blast lifecycle drives capacity counter', () => {
+    it('mitra accept → capacity++; session end is covered separately', async () => {
+      const c = await createCustomer({ callName: 'BlastC' })
+      const m = await createMitra({ callName: 'BlastM', isOnline: true })
+
+      const pay = await createPaymentSession({
+        customerId: c.id,
+        durationMinutes: 15,
+        amount: 30000,
+      })
+      await confirmPaymentSession(pay.id, c.id)
+      const session = await createPairingRequest(c.id, { paymentRequestId: pay.id })
+      expect(await v().get(vkCapacityKey(m.id))).toBe('0') // no accept yet
+
+      await acceptPairingRequest(session.id, m.id)
+      expect(await v().get(vkCapacityKey(m.id))).toBe('1')
+    })
+  })
+
+  // ---------- Reader fallback when Valkey is unavailable ----------
+
+  describe('reader fallback', () => {
+    it('isMitraReachable falls back to Postgres on Valkey error', async () => {
+      const m = await createMitra({ callName: 'Fallback', isOnline: true })
+      // Stub sismember to throw
+      const spy = vi.spyOn(valkey, 'sismember').mockRejectedValue(new Error('valkey down'))
+      try {
+        // Postgres has is_online=true → fallback returns true
+        const result = await isMitraReachable(m.id)
+        expect(result).toBe(true)
+      } finally {
+        spy.mockRestore()
+      }
+    })
+
+    it('findAvailableMitras falls back to Postgres JOIN when Valkey sdiff throws', async () => {
+      const m = await createMitra({ callName: 'FallbackBlast', isOnline: true })
+      const spy = vi.spyOn(valkey, 'sdiff').mockRejectedValue(new Error('valkey down'))
+      try {
+        const result = await findAvailableMitras()
+        expect(result.find(r => r.id === m.id)).toBeDefined()
+      } finally {
+        spy.mockRestore()
+      }
+    })
+  })
+})
diff --git a/backend/test/services/session-timer.service.test.js b/backend/test/services/session-timer.service.test.js
index 46f199b..6489848 100644
--- a/backend/test/services/session-timer.service.test.js
+++ b/backend/test/services/session-timer.service.test.js
@@ -1,4 +1,5 @@
 import { describe, it, expect, beforeEach, afterEach, vi } from 'vitest'
+import { randomUUID } from 'node:crypto'
 
 // Capture calls to sendToSessionParticipant so we can assert the 3-min warning event.
 vi.mock('../../src/plugins/websocket.js', () => ({
@@ -15,10 +16,42 @@ vi.mock('../../src/services/notification.service.js', () => ({
   registerDeviceToken: vi.fn(async () => {}),
 }))
 
-vi.mock('../../src/plugins/valkey.js', () => ({
-  publish: vi.fn(async () => {}),
-  subscribe: vi.fn(() => () => {}),
-}))
+// Real DB queries don't settle under fake timers (they're real socket I/O, not
+// microtasks). Stub getDb() with a tagged-template-compatible mock so onThreeMinuteWarning's
+// `SELECT expires_at FROM chat_sessions WHERE id = ${sessionId}` resolves synchronously.
+vi.mock('../../src/db/client.js', () => {
+  const fakeSql = () => Promise.resolve([{ expires_at: null }])
+  fakeSql.unsafe = () => Promise.resolve([])
+  fakeSql.array = (arr) => arr
+  fakeSql.json = (v) => v
+  return { getDb: () => fakeSql }
+})
+
+vi.mock('../../src/plugins/valkey.js', () => {
+  const noopPipeline = { sadd: () => noopPipeline, srem: () => noopPipeline, set: () => noopPipeline, get: () => noopPipeline, del: () => noopPipeline, exec: async () => [] }
+  return {
+    publish: vi.fn(async () => {}),
+    subscribe: vi.fn(() => () => {}),
+    onValkeyReady: vi.fn(),
+    getValkeyClient: vi.fn(() => ({ setex: vi.fn(async () => 'OK') })),
+    getValkeyPub: vi.fn(),
+    getValkeySub: vi.fn(),
+    sadd: vi.fn(async () => 1),
+    srem: vi.fn(async () => 1),
+    sismember: vi.fn(async () => false),
+    smembers: vi.fn(async () => []),
+    sdiff: vi.fn(async () => []),
+    scard: vi.fn(async () => 0),
+    set: vi.fn(async () => 'OK'),
+    get: vi.fn(async () => null),
+    del: vi.fn(async () => 1),
+    incr: vi.fn(async () => 1),
+    decr: vi.fn(async () => 0),
+    exists: vi.fn(async () => 0),
+    pipeline: vi.fn(() => noopPipeline),
+    multi: vi.fn(() => noopPipeline),
+  }
+})
 
 const { sendToSessionParticipant } = await import('../../src/plugins/websocket.js')
 const { startSessionTimer, clearSessionTimer } = await import('../../src/services/session-timer.service.js')
@@ -35,7 +68,9 @@ describe('session-timer 3-minute warning (Phase 4)', () => {
   })
 
   it('emits session_warning kind:three_minutes_left exactly once at the 3-min mark', async () => {
-    const sessionId = 'sess-3min-test'
+    // Real UUID — onThreeMinuteWarning runs a Postgres SELECT against chat_sessions.id
+    // which is uuid-typed; string ids throw a parse error before we hit the row check.
+    const sessionId = randomUUID()
     const expiresAt = new Date(Date.now() + 5 * 60_000) // 5 minutes from now
 
     startSessionTimer(sessionId, expiresAt)
@@ -65,7 +100,7 @@ describe('session-timer 3-minute warning (Phase 4)', () => {
   })
 
   it('does NOT re-fire the 3-min warning when the timer is rescheduled (e.g. extension)', async () => {
-    const sessionId = 'sess-rescheduled'
+    const sessionId = randomUUID()
     const initial = new Date(Date.now() + 5 * 60_000)
     startSessionTimer(sessionId, initial)
 
diff --git a/requirement/deployment.md b/requirement/deployment.md
new file mode 100644
index 0000000..df18cb9
--- /dev/null
+++ b/requirement/deployment.md
@@ -0,0 +1,107 @@
+# Deployment notes
+
+Operational decisions and dependency configuration for staging/production. Keep this updated as we make infra choices; cross-link from feature plans when a deploy-time setting matters.
+
+## Infrastructure summary
+
+| Component | Service | Tier / Notes |
+|---|---|---|
+| Backend (public + internal) | GCP Cloud Run | Horizontal scaling; SIGTERM trapped for graceful drain ([server.js](../backend/src/server.js)) |
+| Database | GCP Cloud SQL (PostgreSQL) | Source of truth for all durable state |
+| Pub/sub + cache | Valkey | Self-hosted on VM today; Memorystore Standard (HA) recommended for prod (see [§ Valkey](#valkey)) |
+| Networking | GCP VPC | Internal listener (port 3001) never exposed; CC reaches it via VPN |
+| Payment | Xendit | See [phase5-xendit-plan.md](phase5-xendit-plan.md) for keys / webhook URL setup |
+| Auth | Self-managed JWT + FCM-only Firebase | See [backend/CLAUDE.md](../backend/CLAUDE.md) |
+
+## Valkey
+
+Valkey is used for two distinct purposes:
+
+1. **Pub/sub** — cross-instance event fan-out (chat messages, session lifecycle, config invalidation). See [backend/src/plugins/valkey.js](../backend/src/plugins/valkey.js).
+2. **Availability mirror** — `mitras:online`, `mitras:deactivated`, `mitra:capacity:<id>`, `mitra:heartbeat:<id>`, and `availability:snapshot` per [valkey-online-mirror-plan.md](valkey-online-mirror-plan.md). Postgres remains the durable source of truth; Valkey is the hot read path.
+
+### Persistence — required or optional?
+
+**Not required.** All durable state lives in Postgres; Valkey is a cache + ephemeral liveness layer that fully rebuilds via `seedFromPostgres()` on backend reconnect.
+
+What's actually in Valkey, and what happens if it's wiped:
+
+| Key | Derivable from Postgres? | Cost of loss |
+|---|---|---|
+| `mitras:online` | yes | reseeded on reconnect |
+| `mitras:deactivated` | yes | reseeded on reconnect |
+| `mitra:capacity:<id>` | yes (`COUNT(*) FROM chat_sessions`) | reseeded on reconnect |
+| `mitra:heartbeat:<id>` | **no** — pure transient liveness | seed writes `NOW`; ≤ a few seconds of fuzz on `last_heartbeat_at` forensics |
+| `availability:snapshot` | recomputable | next beacon poll repopulates |
+
+Reader code in services/* has explicit Postgres fallbacks for every Valkey op, so the cold-cache window during a restart degrades performance, not correctness.
+
+### Persistence recommendation by environment
+
+| Environment | Setting | Reason |
+|---|---|---|
+| **Dev / local** | No persistence (`--save "" --appendonly no` or just default) | Restarts wipe state; reseed handles it cleanly; zero disk overhead |
+| **Staging** | AOF on (`--appendonly yes`) | Verifies prod-like behavior; tiny disk cost |
+| **Production** | AOF on, optionally RDB too (`--appendonly yes --save 60 1000`) | Eliminates cold-cache window after restart; trivial disk footprint (few MB) |
+
+The application code is identical across all three — persistence is a deploy-time knob, not a code-level concern.
+
+### Self-hosted Valkey (current state, dev/staging)
+
+Docker container on the existing VM. Reference config:
+
+```yaml
+valkey:
+  image: valkey/valkey:7-alpine
+  command: valkey-server --appendonly yes --save 60 1000
+  volumes:
+    - valkey-data:/data
+  ports:
+    - "6379:6379"
+  restart: unless-stopped
+```
+
+Backend reaches it via `VALKEY_URL=redis://<vm-ip>:6379` in `backend/.env` (or Cloud Run env var).
+
+### Memorystore migration (when going to prod)
+
+The reseed-from-Postgres flow makes migration trivial — Valkey state is never load-bearing:
+
+1. Provision **Memorystore for Valkey, Standard tier** (HA with replica) in the same VPC + region as Cloud Run.
+   - Smallest available size (~1 GB) is plenty; actual data footprint is well under 1 MB.
+   - Cost: ~$50/month at minimum sizing in asia-southeast2.
+2. Update Cloud Run env: `VALKEY_URL=redis://<memorystore-internal-ip>:6379`.
+3. Deploy new revision. Cloud Run rolling deploy → new instances seed Memorystore from Postgres; old instances drain on old Valkey.
+4. Shut down old Valkey once traffic has migrated.
+
+**Zero downtime.** No data migration needed (state is derivable). The cold-cache window on new instances is handled by the existing Postgres-fallback reader paths.
+
+### Tier choice rationale
+
+| Tier | When to use | Failover behavior |
+|---|---|---|
+| Self-hosted Docker | Dev, staging | Manual restart; backend reseeds when Valkey comes back |
+| Memorystore Basic | Cost-sensitive single-AZ staging | ~1–5 min outage per maintenance event; backend handles via Postgres fallback |
+| Memorystore Standard (HA) | **Production** | ~30s automatic failover; replica keeps data live |
+
+The system is correct on any tier — HA reduces customer-visible latency spikes during Valkey events from minutes to seconds.
+
+## Cloud Run
+
+(Placeholder — fill in as we make decisions about region, min/max instances, concurrency, secrets manager wiring.)
+
+## Cloud SQL
+
+(Placeholder — pool size, machine type, HA flag, backup retention.)
+
+## Xendit
+
+See [phase5-xendit-plan.md](phase5-xendit-plan.md) for credential setup and webhook URL configuration. Stage 8 (live E2E) is currently blocked on test-mode keys.
+
+## Open ops decisions
+
+- [ ] Confirm Memorystore Standard tier for prod deploy (recommended in [§ Valkey](#valkey)).
+- [ ] Pin GCP region for backend + Cloud SQL + Memorystore (all must match for sub-ms internal latency).
+- [ ] Secrets manager (GCP Secret Manager vs Cloud Run env vars) for `AUTH_JWT_SECRET`, `XENDIT_SECRET_KEY`, etc.
+- [ ] Backup retention policy for Cloud SQL.
+- [ ] CI/CD pipeline for Cloud Run deploys.
diff --git a/requirement/valkey-online-mirror-plan.md b/requirement/valkey-online-mirror-plan.md
new file mode 100644
index 0000000..1d1040c
--- /dev/null
+++ b/requirement/valkey-online-mirror-plan.md
@@ -0,0 +1,376 @@
+# Valkey mirror for mitra availability — plan
+
+**Status:** approved (2026-05-25). Open issue: heartbeat auth preHandler (see [§ Open issues](#open-issues)).
+**Created:** 2026-05-25
+**Owner:** Ramadhan
+
+## Goal
+
+Move the **read path** for "is this mitra available to blast?" entirely into Valkey at production scale (hundreds of online mitras, thousands of customers polling). Eliminate per-heartbeat Postgres writes. Keep Postgres as the durable source of truth via either real-time mirroring (writes that already exist) or periodic batched mirroring (heartbeats).
+
+## North star
+
+- **Postgres = durable source of truth.** Every fact lives there.
+- **Valkey = read path + ephemeral-write target.** Mirrors Postgres state; reads compute availability from primitive signals.
+- **Valkey unreachable on a read:** fall back to the existing Postgres JOIN query. Outage degrades performance, never breaks pairing.
+- **Valkey unreachable on a write:** log + continue. Reconciliation sweep heals drift.
+- **Valkey restart / cold start:** reseed from Postgres before serving traffic.
+
+## Schema
+
+Four Valkey structures. None has a TTL (heartbeat freshness is computed by sweep, not by Redis expiry):
+
+| Key | Type | Value / members | Updated by |
+|---|---|---|---|
+| `mitras:online` | SET | mitra UUIDs | `setOnline` SADD; `setOffline` SREM; sweep SREM; bulk-SREM on deactivate |
+| `mitras:deactivated` | SET | mitra UUIDs | `updateMitraStatus(is_active=false)` SADD; `updateMitraStatus(is_active=true)` SREM |
+| `mitra:capacity:<id>` | STRING (integer) | count of active+pending_payment sessions assigned to this mitra | INCR on session accept; DECR on session end/expire/cancel; DECR/INCR pair on reroute |
+| `mitra:heartbeat:<id>` | STRING | ISO 8601 timestamp of last heartbeat | heartbeat handler `SET`; `setOnline` `SET` (seed); `setOffline` `DEL` |
+
+Postgres mirror columns (durable):
+- `mitra_online_status.is_online` — written by `setOnline` / `setOffline` / sweep. Already in schema.
+- `mitra_online_status.last_heartbeat_at` — written by the **batched heartbeat mirror** every 60s (NOT per-ping). Already in schema.
+- `mitras.is_active` — written by `updateMitraStatus`. Already in schema.
+- `chat_sessions.status` + `mitra_id` — already source-of-truth for session counts.
+
+## Read path (computing "available for blast")
+
+All reads compute on the fly from Valkey primitives. No memoized `mitras:available` set.
+
+```js
+const findAvailableMitras = async () => {
+  // 1. online minus deactivated
+  const candidates = await valkey.sdiff('mitras:online', 'mitras:deactivated')
+  if (!candidates.length) return []
+
+  // 2. fetch capacity + heartbeat for each candidate (pipelined: 1 roundtrip)
+  const pipe = valkey.pipeline()
+  for (const id of candidates) {
+    pipe.get(`mitra:capacity:${id}`)
+    pipe.get(`mitra:heartbeat:${id}`)
+  }
+  const results = await pipe.exec()
+
+  // 3. filter by capacity + heartbeat freshness
+  const { max_customers_per_mitra } = await getMaxCustomersPerMitra()
+  const { stale_after_seconds } = await getMitraPingConfig()
+  const cutoff = Date.now() - stale_after_seconds * 1000
+
+  const eligible = []
+  for (let i = 0; i < candidates.length; i++) {
+    const count = Number(results[i * 2][1] ?? 0)
+    const heartbeatAt = results[i * 2 + 1][1]
+    if (count >= max_customers_per_mitra) continue
+    if (!heartbeatAt || Date.parse(heartbeatAt) < cutoff) continue
+    eligible.push({ id: candidates[i], active_session_count: count })
+  }
+  return eligible
+}
+```
+
+**Cost at prod scale (300 online):** 1 `SDIFF` + 600 `GET` (pipelined) = ~1ms. Negligible.
+
+### Read sites — what changes
+
+| Caller | Today | After |
+|---|---|---|
+| `isMitraReachable(mitraId)` ([mitra-status.service.js:215](../backend/src/services/mitra-status.service.js#L215)) | `SELECT is_online ...` | `SISMEMBER mitras:online` + check `mitra:heartbeat:<id>` freshness |
+| `findAvailableMitras` ([pairing.service.js:79](../backend/src/services/pairing.service.js#L79)) | full JOIN with chat_sessions | Valkey-driven (above) |
+| `countAvailableMitrasFromCache` ([mitra-status.service.js:181](../backend/src/services/mitra-status.service.js#L181)) | full JOIN, cached in-process 10s | Valkey-driven, cached in **Valkey** 10s (shared cluster-wide) |
+| Dashboard online count ([dashboard.service.js:16](../backend/src/services/dashboard.service.js#L16)) | `COUNT(*) WHERE is_online=true` | `SCARD mitras:online` |
+| `getStatus(mitraId)` (mitra's own status poll) | full SELECT | Hybrid: `SISMEMBER` for `is_online`, Postgres for timestamps |
+| `getOnlineMitras` (CC dashboard) | full JOIN with display_name + active_session_count | **unchanged** — low volume, joins make sense in SQL |
+
+### Reader fallback
+
+Every Valkey read is wrapped:
+
+```js
+try {
+  return await valkey.sismember('mitras:online', mitraId)
+} catch (err) {
+  log.warn({ err }, '[mitras:online] valkey unreachable, falling back to DB')
+  const [row] = await sql`SELECT is_online FROM mitra_online_status WHERE mitra_id = ${mitraId}`
+  return Boolean(row?.is_online)
+}
+```
+
+For `findAvailableMitras`, the fallback is the existing Postgres JOIN query.
+
+## Write paths
+
+Each Postgres write is followed by a Valkey write. Postgres always commits first. Valkey failures log + continue.
+
+### Online / offline (mitra app toggle)
+
+| Action | Postgres | Valkey |
+|---|---|---|
+| `setOnline(mitraId)` | `UPDATE mitra_online_status SET is_online=true, last_online_at=NOW(), last_heartbeat_at=NOW()` | `SADD mitras:online <id>` + `SET mitra:heartbeat:<id> <now-iso>` |
+| `setOffline(mitraId)` | `UPDATE mitra_online_status SET is_online=false, last_offline_at=NOW()` | `SREM mitras:online <id>` + `DEL mitra:heartbeat:<id>` |
+
+The `SET heartbeat` on `setOnline` is critical: without it, the very next sweep would mark the freshly-online mitra stale (their first real heartbeat is up to `heartbeat_cadence_seconds` away).
+
+### Admin activate / deactivate (CC)
+
+[`updateMitraStatus`](../backend/src/services/mitra.service.js) ([mitra.service.js:32](../backend/src/services/mitra.service.js#L32)):
+
+| Action | Postgres | Valkey |
+|---|---|---|
+| activate (`is_active=true`) | `UPDATE mitras SET is_active=true` | `SREM mitras:deactivated <id>` |
+| deactivate (`is_active=false`) | `UPDATE mitras SET is_active=false` | `SADD mitras:deactivated <id>` |
+
+### Mitra heartbeat (hot path)
+
+Heartbeat handler ([mitra-status.service.js:79](../backend/src/services/mitra-status.service.js#L79)) is rewritten to operate entirely against Valkey:
+
+1. `authenticate` plugin verifies JWT signature + expiry + `userType === 'mitra'` (no DB).
+2. `SISMEMBER mitras:deactivated <userId>` → if true, return `403`.
+3. `SET mitra:heartbeat:<userId> <now-iso>`.
+
+Steps 2 + 3 pipelined into one Valkey roundtrip.
+
+| | Today | After |
+|---|---|---|
+| Auth check | `getMitraById` SELECT + `is_active` check (1 DB read) | `SISMEMBER mitras:deactivated` (Valkey only) |
+| Body | `UPDATE mitra_online_status SET last_heartbeat_at=NOW() WHERE id=? AND is_online=true` (1 DB write) | `SET mitra:heartbeat:<id>` (Valkey only) |
+| Postgres ops per ping | 2 | **0** |
+| Valkey ops per ping | 0 | 2 (one pipelined roundtrip) |
+| At prod scale (300 online × 2 pings/min) | 1,200 DB ops/min | 1,200 Valkey ops/min, **0 DB ops/min** |
+
+`mitras:deactivated` is already maintained on every CC `updateMitraStatus` call (see [§ Admin activate / deactivate](#admin-activate--deactivate-cc)) so deactivation propagates to the heartbeat path within one Valkey write window (~ms).
+
+### Heartbeat → Postgres batched mirror
+
+A background job runs every `HEARTBEAT_MIRROR_INTERVAL_SECONDS` (env, default 60). One SQL statement, touches all online mitras at once:
+
+```js
+const mirrorHeartbeatsToPostgres = async () => {
+  const onlineIds = await valkey.smembers('mitras:online')
+  if (!onlineIds.length) return
+
+  const pipe = valkey.pipeline()
+  for (const id of onlineIds) pipe.get(`mitra:heartbeat:${id}`)
+  const results = await pipe.exec()
+
+  const pairs = []
+  for (let i = 0; i < onlineIds.length; i++) {
+    const ts = results[i][1]
+    if (ts) pairs.push({ mitra_id: onlineIds[i], ts })
+  }
+  if (!pairs.length) return
+
+  await sql`
+    UPDATE mitra_online_status m
+    SET last_heartbeat_at = u.ts::timestamptz, updated_at = NOW()
+    FROM (SELECT * FROM UNNEST(
+      ${sql.array(pairs.map(p => p.mitra_id))}::uuid[],
+      ${sql.array(pairs.map(p => p.ts))}::text[]
+    ) AS t(mitra_id, ts)) u
+    WHERE m.mitra_id = u.mitra_id
+  `
+}
+```
+
+**Cost at prod scale (300 online):** 1 batched UPDATE per minute per instance. At 3 instances = 3 statements/min cluster-wide (redundant but idempotent — latest timestamp wins). ~20–60× reduction in Postgres write load vs today.
+
+**No leader election initially.** 3 redundant idempotent UPDATEs/min is still 200× cheaper than today. If WAL pressure becomes visible, add a Valkey-NX lease leader-elect (~15 LOC); deferred.
+
+### Session capacity counter
+
+Touch all four services that mutate session state:
+
+| Trigger | File | Postgres write (existing) | Valkey write |
+|---|---|---|---|
+| Mitra accepts a chat (status → `active`, mitra_id set) | `pairing.service.js` accept handler | `UPDATE chat_sessions SET status='active', mitra_id=?` | `INCR mitra:capacity:<mitra_id>` |
+| Session ends (status → `ended` / `expired` / `cancelled`) | `closure.service.js`, expiry sweepers | `UPDATE chat_sessions SET status=...` | `DECR mitra:capacity:<mitra_id>` |
+| Session reroute (mitra A → mitra B) | `session.service.js` | `UPDATE chat_sessions SET mitra_id=B` | `DECR mitra:capacity:A` + `INCR mitra:capacity:B` (pipelined) |
+
+**What counts as occupying capacity:** sessions in `ACTIVE` or `PENDING_PAYMENT` status with a non-null `mitra_id`. This matches the existing `findAvailableMitras` predicate. Extension flow (active → pending_payment → active) does NOT change capacity — mitra stays occupied throughout.
+
+**Negative-counter guard:** wrap `DECR` in a `Math.max(0, ...)` check via Lua or via a `GET` + `SET` if zero — to prevent drift if a DECR fires without a prior INCR (e.g. legacy session without Valkey mirror). The reconciliation sweep recomputes from Postgres anyway.
+
+## Auto-offline sweep — Valkey-driven
+
+Replaces the current Postgres seq-scan with a Valkey computation. Runs every `MITRA_AUTO_OFFLINE_SWEEP_SECONDS` (env, default 30):
+
+```js
+const autoOfflineStaleMitras = async () => {
+  const { stale_after_seconds, require_ping } = await getMitraPingConfig()
+  if (!require_ping) return 0
+
+  const onlineIds = await valkey.smembers('mitras:online')
+  if (!onlineIds.length) return 0
+
+  const pipe = valkey.pipeline()
+  for (const id of onlineIds) pipe.get(`mitra:heartbeat:${id}`)
+  const results = await pipe.exec()
+
+  const cutoff = Date.now() - stale_after_seconds * 1000
+  const stale = []
+  for (let i = 0; i < onlineIds.length; i++) {
+    const ts = results[i][1]
+    if (!ts || Date.parse(ts) < cutoff) stale.push(onlineIds[i])
+  }
+  if (!stale.length) return 0
+
+  // Postgres: bulk flip is_online=false
+  await sql`
+    UPDATE mitra_online_status
+    SET is_online = false, last_offline_at = NOW(), updated_at = NOW()
+    WHERE mitra_id = ANY(${sql.array(stale)}::uuid[]) AND is_online = true
+  `
+  // Log rows
+  for (const id of stale) {
+    await sql`INSERT INTO mitra_online_logs (mitra_id, status) VALUES (${id}, 'offline')`
+  }
+  // Valkey: bulk SREM + DEL heartbeat keys
+  const cleanup = valkey.pipeline()
+  cleanup.srem('mitras:online', ...stale)
+  for (const id of stale) cleanup.del(`mitra:heartbeat:${id}`)
+  await cleanup.exec()
+
+  invalidateAvailabilityCache()
+  return stale.length
+}
+```
+
+**Stale threshold:** `stale_after_seconds` is read from `getMitraPingConfig()` ([config.service.js](../backend/src/services/config.service.js)) — the existing CC-tunable knob. Not a new env.
+
+**Sweep cadence:** `MITRA_AUTO_OFFLINE_SWEEP_SECONDS` env, default 30 (matches current hardcoded setInterval).
+
+**Failure semantics:** if any Valkey op throws, the entire sweep aborts for this tick. The next tick retries. We never want to mass-offline mitras due to a transient Valkey hiccup.
+
+## Shared beacon snapshot cache
+
+Replace the in-process `availabilityCache` ([mitra-status.service.js:14](../backend/src/services/mitra-status.service.js#L14)) with a Valkey GET/SETEX key. Even though reads are now sub-ms Valkey ops, this caps total Valkey query load at high beacon-poll rates (5,000 customers × 12/min = 60,000 polls/min → cache → 6 computations/min cluster-wide).
+
+```
+KEY:     availability:snapshot
+TYPE:    string (JSON: {"available": bool, "count": number})
+TTL:     10 seconds
+```
+
+`config:invalidate` pub/sub subscriber extended to `DEL availability:snapshot` on `max_customers_per_mitra` bust.
+
+## Startup / reconnect / reseed
+
+Three triggers reseed Valkey state from Postgres. All idempotent.
+
+### Backend startup
+
+In [`server.js`](../backend/src/server.js), after Valkey emits `'ready'` and before the public listener binds:
+
+```js
+const onlineRows = await sql`SELECT mitra_id FROM mitra_online_status WHERE is_online = true`
+const deactRows = await sql`SELECT id FROM mitras WHERE is_active = false`
+const sessionCountRows = await sql`
+  SELECT mitra_id, COUNT(*)::int AS c FROM chat_sessions
+  WHERE mitra_id IS NOT NULL AND status IN ('active', 'pending_payment')
+  GROUP BY mitra_id
+`
+
+const pipe = valkey.multi()
+pipe.del('mitras:online', 'mitras:deactivated')
+if (onlineRows.length) {
+  pipe.sadd('mitras:online', ...onlineRows.map(r => r.mitra_id))
+  // Seed heartbeat timestamps to NOW so the first sweep doesn't mass-offline
+  // currently-online mitras. They'll refresh on their next ping anyway.
+  const now = new Date().toISOString()
+  for (const r of onlineRows) pipe.set(`mitra:heartbeat:${r.mitra_id}`, now)
+}
+if (deactRows.length) pipe.sadd('mitras:deactivated', ...deactRows.map(r => r.id))
+for (const r of sessionCountRows) pipe.set(`mitra:capacity:${r.mitra_id}`, r.c)
+await pipe.exec()
+```
+
+### ioredis reconnect
+
+Listen to the ioredis `'ready'` event (fires on initial connect AND each reconnect). Re-run the seed.
+
+### Periodic reconciliation sweeper
+
+`VALKEY_ONLINE_MIRROR_SWEEP_SECONDS` env, default 300. Runs the seed (idempotent — `DEL` + `SADD` + `SET` lands the same state). Belt-and-braces against drift from failed best-effort writes, out-of-band Postgres mutations, Valkey eviction.
+
+## Two sweeps, two cadences (summary)
+
+| Sweep | Purpose | Cadence env | Default | Reads from | Writes to |
+|---|---|---|---|---|---|
+| Auto-offline | Detect stale-heartbeat mitras → flip offline | `MITRA_AUTO_OFFLINE_SWEEP_SECONDS` | 30 | Valkey | Postgres + Valkey |
+| Heartbeat mirror | Persist Valkey heartbeats to Postgres for forensics/backup | `HEARTBEAT_MIRROR_INTERVAL_SECONDS` | 60 | Valkey | Postgres |
+| Reconciliation | Heal Valkey/Postgres drift | `VALKEY_ONLINE_MIRROR_SWEEP_SECONDS` | 300 | Postgres | Valkey |
+
+All three run on every backend instance independently. All idempotent. No leader election required.
+
+## Multi-instance safety
+
+Cloud Run runs N instances. Each instance:
+
+- Writes both stores on its own mutations. Atomic Valkey ops (`SADD` / `SREM` / `INCR` / `DECR` / `SET`) — no cross-instance coordination needed.
+- Runs all three sweeps independently. Redundant but idempotent.
+- Recompute-on-read for blast eligibility — no stale aggregate to invalidate.
+
+The `mitra:capacity:<id>` counter is the most race-sensitive: `INCR` / `DECR` are atomic but a session-state change must consistently fire exactly one increment and one decrement over its lifetime. The reconciliation sweep recomputes from `chat_sessions` and resets the counter, healing any drift.
+
+## Failure mode summary
+
+| | Behavior |
+|---|---|
+| Valkey unreachable on read | Fall back to Postgres JOIN query |
+| Valkey unreachable on write | Log + continue. Reconciliation sweep heals later. |
+| Postgres unreachable on heartbeat mirror | Skip this cycle. Next cycle writes the latest. |
+| Auto-offline sweep can't reach Valkey | Skip this tick. Mitras stay "online" until Valkey comes back + heartbeat ages out. |
+| Valkey crash (catastrophic) | Backend reconnects → reseed from Postgres. Worst case: ≤60s of `last_heartbeat_at` forensics lost. |
+| Backend crash | Other instances keep running. New instance reseed on startup. |
+
+## Files touched
+
+| File | Change |
+|---|---|
+| `backend/src/plugins/valkey.js` | Add wrappers: `sadd`, `srem`, `sismember`, `smembers`, `sdiff`, `scard`, `set`, `get`, `del`, `incr`, `decr`, `pipeline`/`multi` + `'ready'` reconnect hook |
+| `backend/src/services/config.service.js` | Add `getMitraAutoOfflineSweepSeconds`, `getHeartbeatMirrorIntervalSeconds`, `getValkeyOnlineMirrorSweepSeconds` getters |
+| `backend/src/services/mitra-status.service.js` | Major rewrite (see [§ Read path](#read-path-computing-available-for-blast) and [§ Write paths](#write-paths)). Add `incrementCapacity`, `decrementCapacity`, `mirrorHeartbeatsToPostgres`, `seedFromPostgres` |
+| `backend/src/services/mitra.service.js` | Wrap `updateMitraStatus` with `SADD`/`SREM mitras:deactivated` |
+| `backend/src/services/pairing.service.js` | Rewrite `findAvailableMitras` as Valkey-driven (Postgres fallback). `INCR mitra:capacity` on session accept. |
+| `backend/src/services/closure.service.js` | `DECR mitra:capacity` on session end/expire/cancel |
+| `backend/src/services/session.service.js` | `DECR` + `INCR` pair on reroute |
+| `backend/src/services/dashboard.service.js` | `SCARD mitras:online` for online count |
+| `backend/src/routes/public/mitra.status.routes.js` | Replace `resolveMitra` on `POST /heartbeat` with a Valkey `SISMEMBER mitras:deactivated` check (keep `resolveMitra` on `/online`, `/offline`, `GET /`) |
+| `backend/src/server.js` | Call `seedFromPostgres` on startup (before listener binds); replace hardcoded 30_000 setInterval with env-driven cadence; register heartbeat mirror + reconciliation sweep intervals |
+| `backend/.env.example` | Document `MITRA_AUTO_OFFLINE_SWEEP_SECONDS`, `HEARTBEAT_MIRROR_INTERVAL_SECONDS`, `VALKEY_ONLINE_MIRROR_SWEEP_SECONDS` |
+| `backend/test/services/mitra-status.service.test.js` | Add tests (see [§ Test plan](#test-plan)) |
+| `backend/test/services/pairing.service.test.js` | Update for Valkey-driven `findAvailableMitras` |
+| `backend/test/helpers/valkey.js` (new if absent) | Test helper for clean-slate Valkey state per test |
+
+**Estimated touch:** ~400 LOC + ~200 LOC tests. ~2 days focused work.
+
+## Test plan
+
+### Unit
+1. Mock Valkey; verify each writer calls Postgres → Valkey in correct order, with the seed-heartbeat-on-setOnline and DEL-on-setOffline.
+2. Verify reader fallback path runs when Valkey ops throw.
+3. Verify auto-offline sweep aborts entirely when Valkey ops throw (does NOT mass-offline via Postgres-only path).
+4. Verify capacity counter never goes negative (Math.max guard).
+
+### Integration (real Valkey + Postgres)
+1. **Seed correctness:** insert N online rows + M deactivated + session counts in DB; run startup seed; verify all four Valkey structures match.
+2. **Heartbeat refresh:** `setHeartbeat()` → verify Valkey value updates; check that the periodic mirror writes Postgres `last_heartbeat_at` within one mirror cycle.
+3. **Auto-offline:** insert online mitra, manually expire heartbeat by setting timestamp in past, run sweep, verify Postgres `is_online=false` + Valkey `SREM` + `DEL` heartbeat key.
+4. **Capacity lifecycle:** simulate session accept → end across multiple mitras; verify counter increments/decrements; verify reroute moves count from A to B atomically.
+5. **Restart resilience:** seed state, simulate Valkey restart (FLUSHALL), trigger reconnect handler, verify all four structures reseed correctly.
+6. **Reconciliation:** corrupt Valkey (random SADD of non-existent mitra, wrong capacity counter, missing entries); run reconciliation sweep; verify convergence to Postgres state.
+7. **Fallback:** disable Valkey mid-test; verify `findAvailableMitras` falls back to Postgres JOIN query and returns sensible results.
+
+### Regression
+- All 90/92 existing backend tests should still pass.
+- Maestro flows for pairing (ts-customer-*) should pass unchanged.
+
+## Decisions (locked 2026-05-25)
+
+1. **`revokeAllAuthSessions(mitraId)` added to `updateMitraStatus`** in the same PR. Bounds the deactivation gap to access-token TTL across all mitra routes (not just heartbeat).
+2. **Prod Valkey: Memorystore for Valkey, Standard tier** (HA with replica, smallest available capacity ~1GB). Built-in replication keeps heartbeat timestamps live across failover. Staging/dev can run Basic tier — the reseed-from-Postgres flow handles cold-cache restarts correctly either way.
+3. **Keep `last_heartbeat_at` column.** Written by the 60s batched mirror; remains available for operator forensics ("when was X last seen?"). Drop only if a future audit confirms no consumer reads it.
+
+## Future phases (deferred)
+
+- **Heartbeat → Valkey TTL with keyspace notifications.** Replace timestamp-comparison sweep with `notify-keyspace-events Ex` → instant detection of expired heartbeats. Requires Memorystore config change + a subscriber. Defer until 30s detection lag is the visible bottleneck.
+- **Leader-elected mirror/sweep.** Use a Valkey-NX lease so only one instance runs each background job. ~15 LOC each. Defer until the redundant work shows up in metrics.
diff --git a/requirement/valkey-online-mirror-testing.md b/requirement/valkey-online-mirror-testing.md
new file mode 100644
index 0000000..a108458
--- /dev/null
+++ b/requirement/valkey-online-mirror-testing.md
@@ -0,0 +1,272 @@
+# Valkey Availability Mirror — Testing Checklist
+
+End-to-end verification for the Valkey-mirror refactor described in [valkey-online-mirror-plan.md](valkey-online-mirror-plan.md).
+
+Cluster labels: **[BE]** backend / curl / SQL / Valkey-cli, **[CC]** control_center, **[M]** mitra_app, **[C]** client_app.
+
+> **Run order:** Section A first (seed + startup) — every subsequent section assumes a fresh seed. Sections B–J are otherwise independent.
+
+---
+
+## Setup
+
+- [ ] Backend running on `192.168.88.247:3000` (public) + `:3001` (internal) — `curl http://192.168.88.247:3000/api/shared/auth-providers` returns 200
+- [ ] Valkey reachable from backend (`VALKEY_URL` matches running instance; `[valkey] subscribed to config:invalidate` appears in backend boot log)
+- [ ] Postgres reachable; backend boot log shows `[valkey-mirror] seed: X online, Y deactivated, Z with active sessions`
+- [ ] At least 3 mitra accounts exist in DB (we need to flip them online/offline/deactivated across tests)
+- [ ] One customer account ready for the blast scenarios
+- [ ] Helpful aliases for verification (run from `backend/`):
+  ```bash
+  alias vk='node --env-file=.env -e'
+  # Then in tests: vk "(async()=>{const v=(await import('./src/plugins/valkey.js')).getValkeyClient();
+  #   console.log(await v.smembers('mitras:online'));process.exit(0)})()"
+  ```
+
+---
+
+## Section A — Seed + Startup
+
+Verifies `seedFromPostgres()` populates Valkey correctly from Postgres truth.
+
+- [ ] **[BE]** Restart backend; log shows one `[valkey-mirror] seed: N online, M deactivated, K with active sessions` line on startup
+- [ ] **[BE]** Counts in the log match Postgres truth:
+  ```sql
+  SELECT
+    (SELECT COUNT(*) FROM mitra_online_status WHERE is_online=true) AS online,
+    (SELECT COUNT(*) FROM mitras WHERE is_active=false) AS deactivated,
+    (SELECT COUNT(DISTINCT mitra_id) FROM chat_sessions
+     WHERE mitra_id IS NOT NULL AND status IN ('active','pending_payment')) AS with_sessions;
+  ```
+- [ ] **[BE]** Valkey contents match — `SMEMBERS mitras:online` returns the same IDs as `SELECT mitra_id FROM mitra_online_status WHERE is_online=true`
+- [ ] **[BE]** Valkey `mitra:heartbeat:<id>` exists for every currently-online mitra, value is a recent ISO timestamp (within seed time)
+- [ ] **[BE]** Valkey `mitra:capacity:<id>` matches `SELECT COUNT(*) FROM chat_sessions WHERE mitra_id=<id> AND status IN ('active','pending_payment')` for every online mitra
+- [ ] **[BE]** `SMEMBERS mitras:deactivated` matches `SELECT id FROM mitras WHERE is_active=false`
+
+### Reconnect re-seed
+
+- [ ] **[BE]** With backend running, `FLUSHDB` on Valkey, wait ~5s for ioredis reconnect, verify a second `[valkey-mirror] seed:` log entry appears
+- [ ] **[BE]** All four Valkey structures are rebuilt and match Postgres again
+
+---
+
+## Section B — Online / Offline toggle (write-through)
+
+Verifies `setOnline` / `setOffline` writes both stores in the right order.
+
+- [ ] **[M]** Mitra taps "online" → backend updates Postgres `mitra_online_status.is_online=true`
+- [ ] **[BE]** `SISMEMBER mitras:online <mitra-id>` returns `1` within 1s of the toggle
+- [ ] **[BE]** `GET mitra:heartbeat:<mitra-id>` returns a fresh ISO timestamp (within seconds of the toggle)
+- [ ] **[M]** Mitra taps "offline"
+- [ ] **[BE]** Postgres `is_online=false`, Valkey `SISMEMBER mitras:online <id>` returns `0`, `GET mitra:heartbeat:<id>` returns `nil`
+- [ ] **[BE]** `mitra_online_logs` has paired `online` / `offline` audit rows
+
+### Valkey-failure best-effort write
+
+- [ ] **[BE]** Stop Valkey, then toggle a mitra online → request succeeds (200), backend log shows `[valkey-mirror] setOnline <id> failed:` but Postgres is updated correctly
+- [ ] **[BE]** Restart Valkey → reconciliation sweep (≤ 300s default) eventually rebuilds the SET to include this mitra
+
+---
+
+## Section C — Heartbeat path
+
+Verifies the rewrite: per-ping = 1 Valkey write, 0 DB writes.
+
+- [ ] **[M]** Mitra online for at least one heartbeat cycle (~30s)
+- [ ] **[BE]** Watch Postgres query log during heartbeat — **no `UPDATE mitra_online_status SET last_heartbeat_at` rows fire on every ping** (only the batched mirror, default every 60s)
+- [ ] **[BE]** `GET mitra:heartbeat:<id>` value advances on each ping (re-check ~30s later, timestamp moves forward)
+- [ ] **[BE]** After 60s+ wait, `SELECT last_heartbeat_at FROM mitra_online_status WHERE mitra_id=<id>` advances (heartbeat mirror sweep ran)
+
+### Deactivation guard via Valkey
+
+- [ ] **[CC]** Admin deactivates the mitra (Phase 5 path: `is_active=false`)
+- [ ] **[BE]** `SISMEMBER mitras:deactivated <id>` immediately returns `1`
+- [ ] **[M]** Mitra app's next heartbeat → backend returns `403 ACCOUNT_INACTIVE`
+- [ ] **[BE]** No Postgres SELECT on `mitras` table for that heartbeat (verify with query log) — guard is pure Valkey
+
+### Fallback when Valkey is down
+
+- [ ] **[BE]** Stop Valkey
+- [ ] **[M]** Mitra app heartbeats → backend logs `[heartbeat] valkey check failed, falling back to DB`, request still succeeds for active mitra, returns `403` for deactivated mitra (full DB-backed resolveMitra path)
+- [ ] **[BE]** Restart Valkey → next heartbeat uses Valkey path again (no fallback log line)
+
+---
+
+## Section D — Capacity counter
+
+Verifies INCR/DECR across session lifecycle via `recomputeCapacityForMitra`.
+
+- [ ] **[BE]** Reset: pick a mitra with `mitra:capacity:<id> = 0`
+- [ ] **[C]** Customer pays → mitra accepts the blast → chat starts
+- [ ] **[BE]** `GET mitra:capacity:<id>` returns `1` within 1s of mitra accepting
+- [ ] **[C]** Second customer pays → same mitra accepts (assuming `max_customers_per_mitra >= 2`)
+- [ ] **[BE]** `GET mitra:capacity:<id>` returns `2`
+- [ ] **[C]** First session ends naturally (timer expires + goodbye flow completes)
+- [ ] **[BE]** `GET mitra:capacity:<id>` returns `1` within 1s
+- [ ] **[C]** Second session ends
+- [ ] **[BE]** `GET mitra:capacity:<id>` returns `0`
+
+### Reroute
+
+- [ ] **[CC]** Reroute an active session from mitra A → mitra B
+- [ ] **[BE]** `mitra:capacity:A` decrements, `mitra:capacity:B` increments — both atomic with the chat_sessions UPDATE
+
+### Capacity gates blast
+
+- [ ] **[BE]** Set `mitra:capacity:<id>` to `max_customers_per_mitra` directly (`SET mitra:capacity:<id> 3`)
+- [ ] **[C]** Customer pays → blast → **this mitra is excluded** (verify with `chat_request_notifications` — no row for this mitra)
+
+---
+
+## Section E — Deactivation flow (full propagation)
+
+Verifies `updateMitraStatus` + `revokeAllSessionsForUser` close the deactivation gap.
+
+- [ ] **[BE]** Mitra has an active access token (capture from `/api/mitra/auth/otp/verify` or use existing logged-in session)
+- [ ] **[BE]** Confirm mitra can call protected routes (`curl -H "Authorization: Bearer <token>" /api/mitra/...`)
+- [ ] **[CC]** Admin deactivates the mitra
+- [ ] **[BE]** Postgres: `mitras.is_active=false`, `auth_sessions.revoked_at IS NOT NULL` for this mitra
+- [ ] **[BE]** Valkey: `SISMEMBER mitras:deactivated <id>` = 1
+- [ ] **[BE]** Mitra's current access token still works on routes that don't re-check active state (stateless JWT) — bounded by access-token TTL
+- [ ] **[BE]** Mitra's heartbeat returns `403 ACCOUNT_INACTIVE` immediately (Valkey check on hot path)
+- [ ] **[BE]** Mitra's next refresh token attempt fails (because `auth_sessions.revoked_at` was set) → app effectively logs them out
+
+### Re-activation
+
+- [ ] **[CC]** Re-activate the mitra
+- [ ] **[BE]** Postgres `is_active=true`, Valkey `SISMEMBER mitras:deactivated <id>` = 0
+- [ ] **[M]** Mitra re-logs in, heartbeats again successfully
+
+---
+
+## Section F — Read paths (Valkey-first)
+
+Verifies all reads use Valkey with Postgres fallback.
+
+### `isMitraReachable`
+
+- [ ] **[BE]** Online mitra with fresh heartbeat → `isMitraReachable` returns `true` (call via `node -e` or any route that uses it, e.g. extension flow)
+- [ ] **[BE]** Manually set `mitra:heartbeat:<id>` to a timestamp older than `stale_after_seconds` → `isMitraReachable` returns `false` (even though `is_online=true` in Postgres)
+- [ ] **[BE]** Stop Valkey → `isMitraReachable` logs `[isMitraReachable] valkey unavailable, falling back to DB` and returns based on Postgres `is_online`
+
+### `findAvailableMitras` (blast)
+
+- [ ] **[BE]** Run `findAvailableMitras` (e.g. trigger a customer blast) — log shows Valkey path used (no warning about fallback)
+- [ ] **[BE]** Result IDs match `SDIFF mitras:online mitras:deactivated` filtered by capacity + heartbeat freshness
+- [ ] **[BE]** Stop Valkey → next blast logs `[findAvailableMitras] valkey unavailable, falling back to Postgres` and still returns correct results
+
+### `countAvailableMitrasFromCache` (customer beacon)
+
+- [ ] **[BE]** `curl /public/bestie-availability` returns `{available: bool, count: N}` matching reality
+- [ ] **[BE]** `GET availability:snapshot` in Valkey shows cached JSON within 10s of last poll
+- [ ] **[BE]** Multiple rapid polls (5+ per second from 3 different IPs) → only one Valkey-driven recompute per 10s; Postgres query log shows **zero** mitra-availability JOINs in steady state (only the once-per-10s cache miss)
+- [ ] **[CC]** Operator changes `max_customers_per_mitra` → `availability:snapshot` is `DEL`d (cache bust), next poll recomputes
+
+### Dashboard online count
+
+- [ ] **[CC]** Dashboard "Online Mitras" stat matches `SCARD mitras:online` in Valkey
+- [ ] **[BE]** Verify the dashboard query no longer hits `SELECT COUNT(*) FROM mitra_online_status WHERE is_online=true` (check query log)
+
+---
+
+## Section G — Auto-offline sweep (Valkey-driven)
+
+Verifies stale heartbeats → flipped offline via Valkey diff.
+
+- [ ] **[BE]** Set `MITRA_AUTO_OFFLINE_SWEEP_SECONDS=10` in env for faster test cycle, restart backend
+- [ ] **[CC]** Set `stale_after_seconds=15` in CC settings (or directly in `app_config`)
+- [ ] **[M]** Mitra goes online, sends one heartbeat
+- [ ] **[BE]** Manually delete the heartbeat key: `DEL mitra:heartbeat:<id>`
+- [ ] **[BE]** Wait up to 25s (15s stale + 10s sweep cadence)
+- [ ] **[BE]** Postgres: `is_online=false` for this mitra, audit row in `mitra_online_logs` with status='offline'
+- [ ] **[BE]** Valkey: `SISMEMBER mitras:online <id>` = 0, no `mitra:heartbeat:<id>` key
+
+### Sweep skips on Valkey error
+
+- [ ] **[BE]** Stop Valkey
+- [ ] **[BE]** Backend log shows `[auto-offline] valkey unavailable, skipping this tick:` each sweep cadence
+- [ ] **[BE]** **No Postgres UPDATE fires** during the outage (verify with query log) — confirms we don't mass-offline on Valkey hiccup
+
+---
+
+## Section H — Heartbeat → Postgres batched mirror
+
+Verifies the 60s UNNEST UPDATE.
+
+- [ ] **[BE]** Multiple mitras online, all heartbeating
+- [ ] **[BE]** Set `HEARTBEAT_MIRROR_INTERVAL_SECONDS=15` for faster cycles, restart
+- [ ] **[BE]** Wait one mirror cycle — Postgres log shows **one** UPDATE statement (with `UNNEST(...)` containing all online mitra IDs)
+- [ ] **[BE]** `SELECT last_heartbeat_at FROM mitra_online_status WHERE is_online=true` returns timestamps within last cycle
+- [ ] **[BE]** Compare with Valkey `GET mitra:heartbeat:<id>` — Postgres lags Valkey by ≤ mirror-cadence seconds (forensic-grade, not real-time)
+
+### Mirror skips on Valkey error
+
+- [ ] **[BE]** Stop Valkey
+- [ ] **[BE]** Backend log shows `[heartbeat-mirror] valkey unavailable, skipping:` each cycle
+- [ ] **[BE]** Postgres `last_heartbeat_at` does NOT advance during the outage (correct — Valkey is the source of "when was last ping?")
+
+---
+
+## Section I — Reconciliation sweep
+
+Verifies drift heals every `VALKEY_ONLINE_MIRROR_SWEEP_SECONDS`.
+
+- [ ] **[BE]** Set `VALKEY_ONLINE_MIRROR_SWEEP_SECONDS=30` for faster test, restart
+- [ ] **[BE]** Manually corrupt Valkey:
+  ```
+  SADD mitras:online 00000000-0000-0000-0000-000000000999  # bogus ID
+  SREM mitras:online <real-online-mitra-id>                 # remove a real one
+  SET mitra:capacity:<real-id> 99                           # bogus capacity
+  ```
+- [ ] **[BE]** Wait one sweep cycle (~30s) → log shows `[valkey-mirror] seed:` again
+- [ ] **[BE]** After sweep: Valkey state matches Postgres exactly (bogus ID gone, real ID present, capacity reset)
+
+### Sweep disabled when env=0
+
+- [ ] **[BE]** Set `VALKEY_ONLINE_MIRROR_SWEEP_SECONDS=0`, restart
+- [ ] **[BE]** Confirm no periodic seed log appears after the initial startup seed
+
+---
+
+## Section J — Failure modes (Valkey degradation)
+
+End-to-end behavior when Valkey is down for an extended period.
+
+- [ ] **[BE]** Stop Valkey
+- [ ] **[C]** Customer beacon poll → `availability:snapshot` GET fails → falls back to Postgres JOIN; UX unchanged but DB query rate spikes
+- [ ] **[C]** Customer triggers blast → `findAvailableMitras` Valkey path fails → falls back to Postgres JOIN; blast still works
+- [ ] **[M]** Mitra heartbeat → Valkey write fails (logged), but heartbeat returns 200; the missed write is irrelevant (auto-offline sweep is also skipping)
+- [ ] **[M]** Mitra toggle online → Postgres update succeeds, Valkey SADD fails (logged); on next reconciliation sweep after Valkey returns, mitra is back in `mitras:online` SET
+- [ ] **[BE]** Restart Valkey → reconnect listener fires → `seedFromPostgres()` runs → state restored; degraded period ends
+
+---
+
+## Section K — Multi-instance (defer until Cloud Run)
+
+Run only when ≥ 2 backend instances are active.
+
+- [ ] **[BE]** Two instances both run their own seed on startup — final Valkey state is consistent (idempotent `DEL + SADD`)
+- [ ] **[BE]** Concurrent setOnline calls on the same mitra from different instances → final SET state correct (atomic SADD)
+- [ ] **[BE]** `availability:snapshot` cache miss on instance A fills the snapshot; instance B's next poll reads the cached value (cluster-shared cache works)
+- [ ] **[BE]** Operator changes `max_customers_per_mitra` on one instance → `config:invalidate` pub/sub fires → other instance also DELs `availability:snapshot`
+- [ ] **[BE]** Heartbeat mirror UPDATEs from multiple instances are idempotent (last writer wins on timestamp, no errors)
+
+---
+
+## Smoke tests (quick happy path)
+
+5-minute sanity check after any deploy:
+
+- [ ] **[BE]** Backend log shows successful seed on startup
+- [ ] **[M]** Mitra toggles online → appears in `SMEMBERS mitras:online`
+- [ ] **[C]** Customer sees "Mulai Curhat" enabled
+- [ ] **[C]** Customer pays → mitra accepts → chat starts → `mitra:capacity:<id>` increments
+- [ ] **[C]** Chat ends → counter decrements
+- [ ] **[M]** Mitra toggles offline → removed from SET
+
+---
+
+## Known limitations / what this checklist does NOT cover
+
+- **Load testing** — sustained heartbeat volume at prod scale (300+ mitras × 2 pings/min). Plan: separate load-test stage when prod is provisioned.
+- **Memorystore-specific behavior** — failover, RDB+AOF interaction. Plan: re-run Sections A, G, I, J against Memorystore Standard tier before prod cutover.
+- **Long-running drift** — overnight runs where eviction or memory-pressure could affect Valkey state. Plan: monitor `INFO memory` in prod for the first week.