feat(bench): self-contained Docker Compose setup

Build benchmark binary inside container via multi-stage
Dockerfile. All-in-one: docker compose run bench.

bench/ClientSim.hs

@@ -0,0 +1,133 @@
+{-# LANGUAGE DataKinds #-}
+{-# LANGUAGE DuplicateRecordFields #-}
+{-# LANGUAGE GADTs #-}
+{-# LANGUAGE LambdaCase #-}
+{-# LANGUAGE NamedFieldPuns #-}
+{-# LANGUAGE OverloadedStrings #-}
+{-# LANGUAGE PatternSynonyms #-}
+{-# LANGUAGE ScopedTypeVariables #-}
+{-# LANGUAGE TupleSections #-}
+{-# LANGUAGE TypeApplications #-}
+{-# OPTIONS_GHC -fno-warn-ambiguous-fields #-}
+
+module ClientSim
+  ( SimClient (..),
+    connectClient,
+    createQueue,
+    subscribeQueue,
+    sendMessage,
+    receiveAndAck,
+    connectN,
+    benchKeyHash,
+  )
+where
+
+import Control.Concurrent.Async (mapConcurrently)
+import Control.Concurrent.STM
+import Control.Monad (forM_)
+import Control.Monad.Except (runExceptT)
+import Data.ByteString.Char8 (ByteString)
+import Data.List (unfoldr)
+import qualified Data.List.NonEmpty as L
+import Network.Socket (ServiceName)
+import qualified Simplex.Messaging.Crypto as C
+import Simplex.Messaging.Encoding
+import Simplex.Messaging.Protocol
+import Simplex.Messaging.Transport
+import Simplex.Messaging.Transport.Client
+import Simplex.Messaging.Version
+
+data SimClient = SimClient
+  { scHandle :: THandleSMP TLS 'TClient,
+    scRcvKey :: C.APrivateAuthKey,
+    scRcvId :: RecipientId,
+    scSndId :: SenderId,
+    scDhSecret :: C.DhSecret 'C.X25519
+  }
+
+benchKeyHash :: C.KeyHash
+benchKeyHash = "LcJUMfVhwD8yxjAiSaDzzGF3-kLG4Uh0Fl_ZIjrRwjI="
+
+connectClient :: TransportHost -> ServiceName -> IO (THandleSMP TLS 'TClient)
+connectClient host port = do
+  let tcConfig = defaultTransportClientConfig {clientALPN = Just alpnSupportedSMPHandshakes}
+  runTransportClient tcConfig Nothing host port (Just benchKeyHash) $ \h ->
+    runExceptT (smpClientHandshake h Nothing benchKeyHash supportedClientSMPRelayVRange False Nothing) >>= \case
+      Right th -> pure th
+      Left e -> error $ "SMP handshake failed: " <> show e
+
+connectN :: Int -> TransportHost -> ServiceName -> IO [THandleSMP TLS 'TClient]
+connectN n host port = do
+  let batches = chunksOf 100 [1 .. n]
+  concat <$> mapM (\batch -> mapConcurrently (\_ -> connectClient host port) batch) batches
+
+createQueue :: THandleSMP TLS 'TClient -> IO SimClient
+createQueue h = do
+  g <- C.newRandom
+  (rPub, rKey) <- atomically $ C.generateAuthKeyPair C.SEd448 g
+  (sPub, sKey) <- atomically $ C.generateAuthKeyPair C.SEd25519 g
+  (dhPub, dhPriv :: C.PrivateKeyX25519) <- atomically $ C.generateKeyPair g
+  -- NEW command
+  Resp "1" NoEntity (Ids rId sId srvDh) <- signSendRecv h rKey ("1", NoEntity, New rPub dhPub)
+  let dhShared = C.dh' srvDh dhPriv
+  -- KEY command (secure queue)
+  Resp "2" _ OK <- signSendRecv h rKey ("2", rId, KEY sPub)
+  pure SimClient {scHandle = h, scRcvKey = rKey, scRcvId = rId, scSndId = sId, scDhSecret = dhShared}
+
+subscribeQueue :: SimClient -> IO ()
+subscribeQueue SimClient {scHandle = h, scRcvKey = rKey, scRcvId = rId} = do
+  Resp "3" _ (SOK _) <- signSendRecv h rKey ("3", rId, SUB)
+  pure ()
+
+sendMessage :: THandleSMP TLS 'TClient -> C.APrivateAuthKey -> SenderId -> ByteString -> IO ()
+sendMessage h sKey sId body = do
+  Resp "4" _ OK <- signSendRecv h sKey ("4", sId, SEND noMsgFlags body)
+  pure ()
+
+receiveAndAck :: SimClient -> IO ()
+receiveAndAck SimClient {scHandle = h, scRcvKey = rKey, scRcvId = rId} = do
+  (_, _, Right (MSG RcvMessage {msgId = mId})) <- tGet1 h
+  Resp "5" _ OK <- signSendRecv h rKey ("5", rId, ACK mId)
+  pure ()
+
+-- Helpers (same patterns as ServerTests.hs)
+
+pattern Resp :: CorrId -> EntityId -> BrokerMsg -> Transmission (Either ErrorType BrokerMsg)
+pattern Resp corrId queueId command <- (corrId, queueId, Right command)
+
+pattern Ids :: RecipientId -> SenderId -> RcvPublicDhKey -> BrokerMsg
+pattern Ids rId sId srvDh <- IDS (QIK rId sId srvDh _ _ Nothing Nothing)
+
+pattern New :: RcvPublicAuthKey -> RcvPublicDhKey -> Command 'Creator
+pattern New rPub dhPub = NEW (NewQueueReq rPub dhPub Nothing SMSubscribe (Just (QRMessaging Nothing)) Nothing)
+
+signSendRecv :: (Transport c, PartyI p) => THandleSMP c 'TClient -> C.APrivateAuthKey -> (ByteString, EntityId, Command p) -> IO (Transmission (Either ErrorType BrokerMsg))
+signSendRecv h pk t = do
+  signSend h pk t
+  (r L.:| _) <- tGetClient h
+  pure r
+
+signSend :: (Transport c, PartyI p) => THandleSMP c 'TClient -> C.APrivateAuthKey -> (ByteString, EntityId, Command p) -> IO ()
+signSend h@THandle {params} (C.APrivateAuthKey a pk) (corrId, qId, cmd) = do
+  let TransmissionForAuth {tForAuth, tToSend} = encodeTransmissionForAuth params (CorrId corrId, qId, cmd)
+      authorize t = (,Nothing) <$> case a of
+        C.SEd25519 -> Just . TASignature . C.ASignature C.SEd25519 $ C.sign' pk t
+        C.SEd448 -> Just . TASignature . C.ASignature C.SEd448 $ C.sign' pk t
+        C.SX25519 -> (\THAuthClient {peerServerPubKey = k} -> TAAuthenticator $ C.cbAuthenticate k pk (C.cbNonce corrId) t) <$> thAuth params
+  Right () <- tPut1 h (authorize tForAuth, tToSend)
+  pure ()
+
+tPut1 :: Transport c => THandle v c 'TClient -> SentRawTransmission -> IO (Either TransportError ())
+tPut1 h t = do
+  rs <- tPut h (Right t L.:| [])
+  case rs of
+    (r : _) -> pure r
+    [] -> error "tPut1: empty result"
+
+tGet1 :: (ProtocolEncoding v err cmd, Transport c) => THandle v c 'TClient -> IO (Transmission (Either err cmd))
+tGet1 h = do
+  (r L.:| _) <- tGetClient h
+  pure r
+
+chunksOf :: Int -> [a] -> [[a]]
+chunksOf n = unfoldr $ \xs -> if null xs then Nothing else Just (splitAt n xs)

bench/Dockerfile

@@ -0,0 +1,25 @@
+FROM haskell:9.6.3 AS build
+
+WORKDIR /src
+
+# Copy cabal file first for dependency caching
+COPY simplexmq.cabal cabal.project* ./
+RUN cabal update && cabal build --only-dependencies -f server_postgres smp-server-bench || true
+
+# Copy full source
+COPY . .
+RUN cabal build -f server_postgres smp-server-bench \
+  && cp $(cabal list-bin -f server_postgres smp-server-bench) /usr/local/bin/smp-server-bench
+
+FROM debian:bookworm-slim
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+      libgmp10 libpq5 libffi8 zlib1g ca-certificates \
+  && rm -rf /var/lib/apt/lists/*
+
+COPY --from=build /usr/local/bin/smp-server-bench /usr/local/bin/smp-server-bench
+COPY tests/fixtures /app/tests/fixtures
+
+WORKDIR /app
+
+ENTRYPOINT ["smp-server-bench"]

bench/Main.hs

@@ -0,0 +1,243 @@
+{-# LANGUAGE CPP #-}
+{-# LANGUAGE DataKinds #-}
+{-# LANGUAGE DuplicateRecordFields #-}
+{-# LANGUAGE GADTs #-}
+{-# LANGUAGE LambdaCase #-}
+{-# LANGUAGE NamedFieldPuns #-}
+{-# LANGUAGE NumericUnderscores #-}
+{-# LANGUAGE OverloadedStrings #-}
+{-# LANGUAGE ScopedTypeVariables #-}
+{-# LANGUAGE TypeApplications #-}
+{-# OPTIONS_GHC -fno-warn-ambiguous-fields #-}
+
+module Main where
+
+import Control.Concurrent (threadDelay)
+import Control.Concurrent.Async (async, cancel, forConcurrently_, mapConcurrently, mapConcurrently_)
+import Control.Concurrent.STM
+import Control.Monad (forever, forM_, void, when)
+import Data.ByteString.Char8 (ByteString)
+import qualified Data.ByteString.Char8 as B
+import Data.IORef
+import Data.List (unfoldr)
+import Data.Time.Clock (getCurrentTime, utctDayTime)
+import Network.Socket (ServiceName)
+import System.Environment (getArgs)
+import System.IO (hFlush, stdout)
+
+import ClientSim
+import Report
+
+import Crypto.Random (ChaChaDRG)
+import qualified Simplex.Messaging.Crypto as C
+import Simplex.Messaging.Server (runSMPServerBlocking)
+import Simplex.Messaging.Server.Env.STM as Env
+import Simplex.Messaging.Server.Expiration (ExpirationConfig (..))
+import Simplex.Messaging.Server.MsgStore.Postgres (PostgresMsgStore)
+import Simplex.Messaging.Server.QueueStore.Postgres.Config (PostgresStoreCfg (..))
+import Simplex.Messaging.Agent.Store.Postgres.Options (DBOpts (..))
+import Simplex.Messaging.Agent.Store.Shared (MigrationConfirmation (..))
+import Simplex.Messaging.Client.Agent (SMPClientAgentConfig (..), defaultSMPClientAgentConfig)
+import Simplex.Messaging.Transport
+import Simplex.Messaging.Transport.Server (ServerCredentials (..), mkTransportServerConfig)
+import Simplex.Messaging.Version
+import UnliftIO.Exception (bracket)
+
+import Control.Logger.Simple (logInfo, withGlobalLogging, LogConfig (..), setLogLevel, LogLevel (..))
+
+data BenchConfig = BenchConfig
+  { numClients :: Int,
+    sustainedMinutes :: Int,
+    pgConnStr :: ByteString,
+    serverPort :: ServiceName,
+    timeSeriesFile :: FilePath
+  }
+
+defaultBenchConfig :: BenchConfig
+defaultBenchConfig =
+  BenchConfig
+    { numClients = 5000,
+      sustainedMinutes = 5,
+      pgConnStr = "postgresql://smp@localhost:15432/smp_bench",
+      serverPort = "15001",
+      timeSeriesFile = "bench-timeseries.csv"
+    }
+
+parseArgs :: IO BenchConfig
+parseArgs = do
+  args <- getArgs
+  pure $ go args defaultBenchConfig
+  where
+    go [] c = c
+    go ("--clients" : n : rest) c = go rest c {numClients = read n}
+    go ("--minutes" : n : rest) c = go rest c {sustainedMinutes = read n}
+    go ("--pg" : s : rest) c = go rest c {pgConnStr = B.pack s}
+    go ("--port" : p : rest) c = go rest c {serverPort = p}
+    go ("--timeseries" : f : rest) c = go rest c {timeSeriesFile = f}
+    go (x : _) _ = error $ "Unknown argument: " <> x
+
+main :: IO ()
+main = withGlobalLogging LogConfig {lc_file = Nothing, lc_stderr = True} $ do
+  setLogLevel LogInfo
+  bc@BenchConfig {numClients, sustainedMinutes, serverPort, timeSeriesFile, pgConnStr} <- parseArgs
+  putStrLn $ "SMP Server Memory Benchmark"
+  putStrLn $ "  clients:  " <> show numClients
+  putStrLn $ "  sustain:  " <> show sustainedMinutes <> " min"
+  putStrLn $ "  pg:       " <> B.unpack pgConnStr
+  putStrLn $ "  port:     " <> serverPort
+  putStrLn ""
+
+  snapshotsRef <- newIORef []
+
+  let snap phase clients = do
+        s <- takeSnapshot phase clients
+        modifyIORef' snapshotsRef (s :)
+        putStrLn $ "  [" <> show phase <> "] live=" <> show (snapLive s `div` (1024 * 1024)) <> "MB large=" <> show (snapLarge s `div` (1024 * 1024)) <> "MB"
+        hFlush stdout
+
+  withBenchServer bc $ do
+    putStrLn "Phase 1: Baseline (no clients)"
+    snap "baseline" 0
+
+    putStrLn $ "Phase 2: Connecting " <> show numClients <> " TLS clients..."
+    handles <- connectN numClients "localhost" serverPort
+    putStrLn $ "  Connected " <> show (length handles) <> " clients"
+    snap "tls_connect" (length handles)
+
+    putStrLn "Phase 3: Creating queues (NEW + KEY)..."
+    simClients <- mapConcurrently createQueue handles
+    putStrLn $ "  Created " <> show (length simClients) <> " queues"
+    snap "queue_create" (length simClients)
+
+    putStrLn "Phase 4: Subscribing (SUB)..."
+    mapConcurrently_ subscribeQueue simClients
+    snap "subscribe" (length simClients)
+
+    -- Pair up clients: first half sends to second half
+    let halfN = length simClients `div` 2
+        senders = take halfN simClients
+        receivers = drop halfN simClients
+        pairs = zip senders receivers
+
+    putStrLn $ "Phase 5: Sending " <> show halfN <> " messages..."
+    g <- C.newRandom
+    forConcurrently_ pairs $ \(sender, receiver) -> do
+      (_, sKey) <- atomically $ C.generateAuthKeyPair C.SEd25519 g
+      sendMessage (scHandle sender) sKey (scSndId receiver) "benchmark test message payload 1234567890"
+    snap "msg_send" (length simClients)
+
+    putStrLn "Phase 6: Receiving and ACKing messages..."
+    forConcurrently_ receivers receiveAndAck
+    snap "msg_recv" (length simClients)
+
+    putStrLn $ "Phase 7: Sustained load (" <> show sustainedMinutes <> " min)..."
+    writeTimeSeriesHeader timeSeriesFile
+    -- Logger thread: snapshot every 10s
+    logger <- async $ forever $ do
+      threadDelay 10_000_000
+      s <- takeSnapshot "sustained" (length simClients)
+      appendTimeSeries timeSeriesFile s
+    -- Worker threads: continuous send/receive
+    let loopDurationUs = sustainedMinutes * 60 * 1_000_000
+    workersDone <- newTVarIO False
+    workers <- async $ do
+      deadline <- (+ loopDurationUs) <$> getMonotonicTimeUs
+      sustainedLoop g pairs deadline
+      atomically $ writeTVar workersDone True
+    -- Wait for workers
+    void $ atomically $ readTVar workersDone >>= \done -> when (not done) retry
+    cancel logger
+    cancel workers
+    snap "sustained_end" (length simClients)
+
+  snapshots <- reverse <$> readIORef snapshotsRef
+  printSummary snapshots
+  putStrLn $ "\nTime-series written to: " <> timeSeriesFile
+
+sustainedLoop :: TVar ChaChaDRG -> [(SimClient, SimClient)] -> Int -> IO ()
+sustainedLoop g pairs deadline = go
+  where
+    go = do
+      now <- getMonotonicTimeUs
+      when (now < deadline) $ do
+        forConcurrently_ pairs $ \(sender, receiver) -> do
+          (_, sKey) <- atomically $ C.generateAuthKeyPair C.SEd25519 g
+          sendMessage (scHandle sender) sKey (scSndId receiver) "sustained load message payload"
+        forConcurrently_ (map snd pairs) receiveAndAck
+        go
+
+getMonotonicTimeUs :: IO Int
+getMonotonicTimeUs = do
+  t <- getCurrentTime
+  pure $ round (utctDayTime t * 1_000_000)
+
+withBenchServer :: BenchConfig -> IO a -> IO a
+withBenchServer BenchConfig {pgConnStr, serverPort} action = do
+  started <- newEmptyTMVarIO
+  let srvCfg = benchServerConfig pgConnStr serverPort
+  bracket
+    (async $ runSMPServerBlocking started srvCfg Nothing)
+    cancel
+    (\_ -> waitForServer started >> action)
+  where
+    waitForServer started = do
+      r <- atomically $ takeTMVar started
+      if r
+        then putStrLn $ "Server started on port " <> serverPort
+        else error "Server failed to start"
+
+benchServerConfig :: ByteString -> ServiceName -> ServerConfig PostgresMsgStore
+benchServerConfig pgConn port =
+  let storeCfg = PostgresStoreCfg
+        { dbOpts = DBOpts {connstr = pgConn, schema = "smp_server", poolSize = 10, createSchema = True},
+          dbStoreLogPath = Nothing,
+          confirmMigrations = MCYesUp,
+          deletedTTL = 86400
+        }
+   in ServerConfig
+        { transports = [(port, transport @TLS, False)],
+          smpHandshakeTimeout = 120_000_000,
+          tbqSize = 128,
+          msgQueueQuota = 128,
+          maxJournalMsgCount = 256,
+          maxJournalStateLines = 16,
+          queueIdBytes = 24,
+          msgIdBytes = 24,
+          serverStoreCfg = SSCDatabase storeCfg,
+          storeNtfsFile = Nothing,
+          allowNewQueues = True,
+          newQueueBasicAuth = Nothing,
+          controlPortUserAuth = Nothing,
+          controlPortAdminAuth = Nothing,
+          dailyBlockQueueQuota = 20,
+          messageExpiration = Just defaultMessageExpiration,
+          expireMessagesOnStart = False,
+          expireMessagesOnSend = False,
+          idleQueueInterval = 14400,
+          notificationExpiration = defaultNtfExpiration,
+          inactiveClientExpiration = Nothing,
+          logStatsInterval = Nothing,
+          logStatsStartTime = 0,
+          serverStatsLogFile = "bench/tmp/stats.log",
+          serverStatsBackupFile = Nothing,
+          prometheusInterval = Nothing,
+          prometheusMetricsFile = "bench/tmp/metrics.txt",
+          pendingENDInterval = 500_000,
+          ntfDeliveryInterval = 200_000,
+          smpCredentials =
+            ServerCredentials
+              { caCertificateFile = Just "tests/fixtures/ca.crt",
+                privateKeyFile = "tests/fixtures/server.key",
+                certificateFile = "tests/fixtures/server.crt"
+              },
+          httpCredentials = Nothing,
+          smpServerVRange = supportedServerSMPRelayVRange,
+          Env.transportConfig = mkTransportServerConfig True (Just alpnSupportedSMPHandshakes) True,
+          controlPort = Nothing,
+          smpAgentCfg = defaultSMPClientAgentConfig {persistErrorInterval = 1},
+          allowSMPProxy = False,
+          serverClientConcurrency = 16,
+          information = Nothing,
+          startOptions = StartOptions {maintenance = False, compactLog = False, logLevel = LogInfo, skipWarnings = True, confirmMigrations = MCYesUp}
+        }
+

bench/Report.hs

@@ -0,0 +1,113 @@
+{-# LANGUAGE NamedFieldPuns #-}
+{-# LANGUAGE NumericUnderscores #-}
+{-# LANGUAGE OverloadedStrings #-}
+
+module Report
+  ( Snapshot (..),
+    takeSnapshot,
+    printSummary,
+    writeTimeSeriesHeader,
+    appendTimeSeries,
+  )
+where
+
+import Control.Concurrent (threadDelay)
+import Data.List (foldl')
+import Data.Text (Text)
+import qualified Data.Text as T
+import qualified Data.Text.IO as T
+import Data.Time.Clock (UTCTime, getCurrentTime)
+import Data.Time.Format.ISO8601 (iso8601Show)
+import Data.Word (Word32, Word64)
+import GHC.Stats (RTSStats (..), GCDetails (..), getRTSStats)
+import System.IO (Handle, IOMode (..), hFlush, hSetBuffering, BufferMode (..), withFile)
+import System.Mem (performMajorGC)
+
+data Snapshot = Snapshot
+  { snapTime :: UTCTime,
+    snapPhase :: Text,
+    snapLive :: Word64,
+    snapHeap :: Word64,
+    snapLarge :: Word64,
+    snapFrag :: Word64,
+    snapGCs :: Word32,
+    snapClients :: Int
+  }
+
+takeSnapshot :: Text -> Int -> IO Snapshot
+takeSnapshot phase clients = do
+  performMajorGC
+  threadDelay 1_000_000
+  rts <- getRTSStats
+  ts <- getCurrentTime
+  let GCDetails {gcdetails_live_bytes, gcdetails_mem_in_use_bytes, gcdetails_large_objects_bytes, gcdetails_block_fragmentation_bytes} = gc rts
+  pure
+    Snapshot
+      { snapTime = ts,
+        snapPhase = phase,
+        snapLive = gcdetails_live_bytes,
+        snapHeap = gcdetails_mem_in_use_bytes,
+        snapLarge = gcdetails_large_objects_bytes,
+        snapFrag = gcdetails_block_fragmentation_bytes,
+        snapGCs = gcs rts,
+        snapClients = clients
+      }
+
+printSummary :: [Snapshot] -> IO ()
+printSummary [] = putStrLn "No snapshots collected."
+printSummary snaps = do
+  putStrLn ""
+  putStrLn hdr
+  putStrLn $ replicate (length hdr) '-'
+  mapM_ printRow (zip (Snapshot {snapLive = 0, snapHeap = 0, snapLarge = 0, snapFrag = 0, snapGCs = 0, snapClients = 0, snapPhase = "", snapTime = snapTime (head snaps)} : snaps) snaps)
+  where
+    hdr = padR 20 "Phase" <> padL 12 "live_MB" <> padL 12 "large_MB" <> padL 12 "frag_MB" <> padL 12 "heap_MB" <> padL 10 "clients" <> padL 14 "d_live_MB" <> padL 14 "d_large_MB" <> padL 14 "KB/client"
+    printRow (prev, cur) =
+      putStrLn $
+        padR 20 (T.unpack $ snapPhase cur)
+          <> padL 12 (showMB $ snapLive cur)
+          <> padL 12 (showMB $ snapLarge cur)
+          <> padL 12 (showMB $ snapFrag cur)
+          <> padL 12 (showMB $ snapHeap cur)
+          <> padL 10 (show $ snapClients cur)
+          <> padL 14 (showDeltaMB (snapLive cur) (snapLive prev))
+          <> padL 14 (showDeltaMB (snapLarge cur) (snapLarge prev))
+          <> padL 14 (perClient cur)
+    showMB w = show (w `div` (1024 * 1024))
+    showDeltaMB a b
+      | a >= b = "+" <> show ((a - b) `div` (1024 * 1024))
+      | otherwise = "-" <> show ((b - a) `div` (1024 * 1024))
+    perClient Snapshot {snapClients, snapLive}
+      | snapClients > 0 = show (snapLive `div` fromIntegral snapClients `div` 1024)
+      | otherwise = "-"
+    padR n s = s <> replicate (max 0 (n - length s)) ' '
+    padL n s = replicate (max 0 (n - length s)) ' ' <> s
+
+csvHeader :: Text
+csvHeader = "timestamp,phase,rts_live,rts_heap,rts_large,rts_frag,rts_gc,clients"
+
+snapshotCsv :: Snapshot -> Text
+snapshotCsv Snapshot {snapTime, snapPhase, snapLive, snapHeap, snapLarge, snapFrag, snapGCs, snapClients} =
+  T.intercalate
+    ","
+    [ T.pack $ iso8601Show snapTime,
+      snapPhase,
+      tshow snapLive,
+      tshow snapHeap,
+      tshow snapLarge,
+      tshow snapFrag,
+      tshow snapGCs,
+      tshow snapClients
+    ]
+
+writeTimeSeriesHeader :: FilePath -> IO ()
+writeTimeSeriesHeader path = T.writeFile path (csvHeader <> "\n")
+
+appendTimeSeries :: FilePath -> Snapshot -> IO ()
+appendTimeSeries path snap =
+  withFile path AppendMode $ \h -> do
+    hSetBuffering h LineBuffering
+    T.hPutStrLn h $ snapshotCsv snap
+
+tshow :: Show a => a -> Text
+tshow = T.pack . show

bench/docker-compose.yml

@@ -0,0 +1,46 @@
+services:
+  postgres:
+    image: postgres:17
+    environment:
+      POSTGRES_USER: smp
+      POSTGRES_DB: smp_bench
+      POSTGRES_HOST_AUTH_METHOD: trust
+    volumes:
+      - ./init.sql:/docker-entrypoint-initdb.d/init.sql
+      - pgdata:/var/lib/postgresql/data
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U smp -d smp_bench"]
+      interval: 2s
+      timeout: 5s
+      retries: 10
+
+  bench:
+    build:
+      context: ..
+      dockerfile: bench/Dockerfile
+    depends_on:
+      postgres:
+        condition: service_healthy
+    environment:
+      BENCH_PG: "postgresql://smp@postgres/smp_bench"
+      BENCH_CLIENTS: "${BENCH_CLIENTS:-5000}"
+      BENCH_MINUTES: "${BENCH_MINUTES:-5}"
+    command:
+      - "--pg"
+      - "postgresql://smp@postgres/smp_bench"
+      - "--clients"
+      - "${BENCH_CLIENTS:-5000}"
+      - "--minutes"
+      - "${BENCH_MINUTES:-5}"
+      - "--timeseries"
+      - "/results/timeseries.csv"
+      - "+RTS"
+      - "-N"
+      - "-A16m"
+      - "-T"
+      - "-RTS"
+    volumes:
+      - ./results:/results
+
+volumes:
+  pgdata:

bench/init.sql

@@ -0,0 +1,2 @@
+CREATE EXTENSION IF NOT EXISTS pgcrypto;
+CREATE SCHEMA IF NOT EXISTS smp_server;

bench/run.sh

@@ -0,0 +1,60 @@
+#!/bin/bash
+set -e
+cd "$(dirname "$0")"
+
+mkdir -p results
+
+reset_db() {
+  docker compose down -v 2>/dev/null || true
+  docker compose up -d --wait postgres
+  echo "PostgreSQL ready."
+}
+
+if [ "$1" = "--compare-rts" ]; then
+  shift
+  docker compose build bench
+  for label_flags in \
+    "default:-N -A16m -T" \
+    "F1.2:-N -A16m -F1.2 -T" \
+    "F1.5:-N -A16m -F1.5 -T" \
+    "A4m:-N -A4m -T" \
+    "A4m-F1.2:-N -A4m -F1.2 -T" \
+    "compact:-N -A16m -c -T" \
+    "nonmoving:-N -A16m -xn -T"; do
+    label="${label_flags%%:*}"
+    flags="${label_flags#*:}"
+    echo ""
+    echo "=========================================="
+    echo "  RTS config: $label ($flags)"
+    echo "=========================================="
+    reset_db
+    docker compose run --rm \
+      -e BENCH_CLIENTS="${BENCH_CLIENTS:-1000}" \
+      -e BENCH_MINUTES="${BENCH_MINUTES:-2}" \
+      bench \
+      --pg "postgresql://smp@postgres/smp_bench" \
+      --clients "${BENCH_CLIENTS:-1000}" \
+      --minutes "${BENCH_MINUTES:-2}" \
+      --timeseries "/results/bench-${label}.csv" \
+      "$@" \
+      +RTS $flags -RTS
+  done
+  echo ""
+  echo "Done. Results in bench/results/"
+elif [ "$1" = "--local" ]; then
+  # Run natively (not in container) — requires local Postgres
+  shift
+  reset_db
+  cabal run smp-server-bench -f server_postgres -- \
+    --pg "postgresql://smp@localhost:15432/smp_bench" \
+    --clients "${BENCH_CLIENTS:-5000}" \
+    --minutes "${BENCH_MINUTES:-5}" \
+    "$@" \
+    +RTS -N -A16m -s -RTS
+else
+  # Run fully in containers
+  reset_db
+  docker compose run --rm bench "$@"
+fi
+
+docker compose down

docs/plans/memory-analysis-results.md

@@ -0,0 +1,100 @@
+## Memory Diagnostics Results
+
+### Data Collection
+
+Server: smp19.simplex.im, PostgreSQL backend, `useCache = False`
+RTS flags: `+RTS -N -A16m -I0.01 -Iw15 -s -RTS` (16 cores)
+
+### Mar 20 Data (1 hour, 07:19-08:19)
+
+```
+Time    rts_live  rts_heap  rts_large  rts_frag   clients  non-large
+07:19    7.5 GB    8.2 GB    5.5 GB    0.03 GB    14,000    2.0 GB
+07:24    6.4 GB   10.8 GB    5.2 GB    3.6 GB     14,806    1.2 GB
+07:29    8.2 GB   10.8 GB    6.5 GB    1.8 GB     15,667    1.7 GB
+07:34   10.0 GB   12.3 GB    7.9 GB    1.4 GB     15,845    2.1 GB
+07:39    6.7 GB   13.0 GB    5.3 GB    5.6 GB     16,589    1.4 GB
+07:44    8.5 GB   13.0 GB    6.7 GB    3.7 GB     16,283    1.8 GB
+07:49    6.5 GB   13.0 GB    5.2 GB    5.8 GB     16,532    1.3 GB
+07:54    6.0 GB   13.0 GB    4.8 GB    6.3 GB     16,636    1.2 GB
+07:59    6.4 GB   13.0 GB    5.1 GB    5.9 GB     16,769    1.3 GB
+08:04    8.3 GB   13.0 GB    6.5 GB    3.9 GB     17,352    1.8 GB
+08:09   10.2 GB   13.0 GB    8.0 GB    1.9 GB     17,053    2.2 GB
+08:14    5.6 GB   13.0 GB    4.5 GB    6.8 GB     17,147    1.1 GB
+08:19    7.6 GB   13.0 GB    6.1 GB    4.6 GB     17,496    1.5 GB
+```
+
+non-large = rts_live - rts_large (normal Haskell heap objects: Maps, TVars, closures)
+
+### Mar 19 Data (5.5 hours, 07:49-13:19)
+
+rts_heap grew from 10.1 GB to 20.7 GB over 5.5 hours.
+Post-GC rts_live floor rose from 5.5 GB to 9.1 GB.
+
+### Findings
+
+**1. Large/pinned objects dominate live data (60-80%)**
+
+`rts_large` = 4.5-8.0 GB out of 5.6-10.2 GB live. These are allocations > ~3KB that go on GHC's large object heap. They oscillate (not growing monotonically), meaning they are being allocated and freed constantly — transient, not leaked.
+
+**2. Fragmentation is the heap growth mechanism**
+
+`rts_heap ≈ rts_live + rts_frag`. The heap grows because pinned/large objects fragment GHC's block allocator. Once GHC expands the heap, it never shrinks. Growth pattern:
+- Large objects allocated → occupy blocks
+- Large objects freed → blocks can't be reused if ANY other object shares the block
+- New allocations need fresh blocks → heap expands
+- Heap never returns memory to OS
+
+**3. Non-large heap data is stable (~1.0-2.2 GB)**
+
+Normal Haskell objects (Maps, TVars, closures, client structures) account for only 1-2 GB. This scales with client count at ~100-130 KB/client and does NOT grow over time.
+
+**4. All tracked data structures are NOT the cause**
+
+- `clientSndQ=0, clientMsgQ=0` — TBQueues empty, no message accumulation
+- `smpQSubs` oscillates ~1.0-1.4M — entries are cleaned up, not leaking
+- `ntfStore` < 2K entries — negligible
+- All proxy agent maps near 0
+- `loadedQ=0` — useCache=False confirmed working
+
+**5. Source of large objects is unclear without heap profiling**
+
+The 4.5-8.0 GB of large objects could come from:
+- PostgreSQL driver (`postgresql-simple`/`libpq`) — pinned ByteStrings for query results
+- TLS library (`tls`) — pinned buffers per connection
+- Network socket I/O — pinned ByteStrings for recv/send
+- SMP protocol message blocks
+
+Cannot distinguish between these without `-hT` heap profiling (which is too expensive for this server).
+
+### Root Cause
+
+**GHC heap fragmentation from constant churn of large/pinned ByteString allocations.**
+
+Not a data structure leak. The live data itself is reasonable (5-10 GB for 15-17K clients). The problem is that GHC's copying GC cannot compact around pinned objects, so the heap grows with fragmentation and never shrinks.
+
+### Mitigation Options
+
+All are RTS flag changes — no rebuild needed, reversible by restart.
+
+**1. `-F1.2`** (reduce GC trigger factor from default 2.0)
+- Triggers major GC when heap reaches 1.2x live data instead of 2x
+- Reclaims fragmented blocks sooner
+- Trade-off: more frequent GC, slightly higher CPU
+- Risk: low — just makes GC run more often
+
+**2. Reduce `-A16m` to `-A4m`** (smaller nursery)
+- More frequent minor GC → short-lived pinned objects freed faster
+- Trade-off: more GC cycles, but each is smaller
+- Risk: low — may actually improve latency by reducing GC pause times
+
+**3. `+RTS -xn`** (nonmoving GC)
+- Designed for pinned-heavy workloads — avoids copying entirely
+- Available since GHC 8.10, improved in 9.x
+- Trade-off: different GC characteristics, less battle-tested
+- Risk: medium — different GC algorithm, should test first
+
+**4. Limit concurrent connections** (application-level)
+- Since large objects scale per-client, fewer clients = less fragmentation
+- Trade-off: reduced capacity
+- Risk: low but impacts users

docs/plans/memory-analysis.md

@@ -0,0 +1,225 @@
+## Root Cause Analysis: SMP Server Memory Growth (23.5GB)
+
+### Environment
+
+- **Server**: smp19.simplex.im, ~21,927 connected clients
+- **Storage**: PostgreSQL backend with `useCache = False`
+- **RTS flags**: `+RTS -N -A16m -I0.01 -Iw15 -s -RTS` (16 cores)
+- **Memory**: 23.5GB RES / 1031GB VIRT (75% of available RAM)
+
+### Log Summary
+
+- **Duration**: ~22 hours (Mar 16 12:12 → Mar 17 10:20)
+- **92,277 proxy connection errors** out of 92,656 total log lines (99.6%)
+- **292 unique failing destination servers**, top offender: `nowhere.moe` (12,875 errors)
+- Only **145 successful proxy connections**
+
+---
+
+### Known Factor: GHC Heap Sizing
+
+With 16 cores and `-A16m`:
+- **Nursery**: 16 × 16MB = **256MB baseline**
+- GHC default major GC threshold = **2× live data** — if live data is 10GB, heap grows to ~20GB before major GC
+- The server is rarely idle with 22K clients, so major GC is deferred despite `-I0.01`
+- This is an amplifier — whatever the actual live data size is, GHC roughly doubles it
+
+---
+
+### Candidate Structures That Could Grow Unboundedly
+
+Analysis of the full codebase identified these structures that either grow without bound or have uncertain cleanup:
+
+#### 1. `SubscribedClients` maps — `Env/STM.hs:378`
+
+Both `subscribers.queueSubscribers` and `ntfSubscribers.queueSubscribers` (and their `serviceSubscribers`) use `SubscribedClients (TMap EntityId (TVar (Maybe (Client s))))`.
+
+Comment at line 376: *"The subscriptions that were made at any point are not removed"*
+
+`deleteSubcribedClient` IS called on disconnect (Server.hs:1112) and DOES call `TM.delete`. But it only deletes if the current stored client matches — if another client already re-subscribed, the old client's disconnect won't remove the entry. This is by design for mobile client continuity, but the net effect on map size over time is unclear without measurement.
+
+#### 2. ProxyAgent's subscription TMaps — `Client/Agent.hs:145-151`
+
+The `SMPClientAgent` has 4 TMaps that accumulate one top-level entry per unique destination server and **never remove** them:
+
+- `activeServiceSubs :: TMap SMPServer (TVar ...)` (line 145)
+- `activeQueueSubs :: TMap SMPServer (TMap QueueId ...)` (line 146)
+- `pendingServiceSubs :: TMap SMPServer (TVar ...)` (line 149)
+- `pendingQueueSubs :: TMap SMPServer (TMap QueueId ...)` (line 150)
+
+Comment at line 262: *"these vars are never removed, they are only added"*
+
+These are only used for the proxy agent (SParty 'Sender), so they grow with each unique destination SMP server proxied to. With 292 unique servers in this log period, these are likely small — but long-running servers may accumulate thousands.
+
+`closeSMPClientAgent` (line 369) does NOT clear these 4 maps.
+
+#### 3. `NtfStore` — `NtfStore.hs:26`
+
+`NtfStore (TMap NotifierId (TVar [MsgNtf]))` — one entry per NotifierId.
+
+`deleteExpiredNtfs` (line 47) filters expired notifications from lists but does **not remove entries with empty lists** from the TMap. Over time, NotifierIds that no longer receive notifications leave zombie `TVar []` entries.
+
+`deleteNtfs` (line 44) does remove the full entry via `TM.lookupDelete` — but only called when a notifier is explicitly deleted.
+
+#### 4. `serviceLocks` in PostgresQueueStore — `Postgres.hs:112,469`
+
+`serviceLocks :: TMap CertFingerprint Lock` — one Lock per unique certificate fingerprint.
+
+`getCreateService` (line 469) calls `withLockMap (serviceLocks st) fp` which calls `getMapLock` (Agent/Client.hs:1029-1032) — this **unconditionally inserts** a Lock into the TMap. There is **no cleanup code** for serviceLocks anywhere. This is NOT guarded by `useCache`.
+
+#### 5. `sentCommands` per proxy client connection — `Client.hs:580`
+
+Each `PClient` has `sentCommands :: TMap CorrId (Request err msg)`. Entries are added per command sent (line 1369) and only removed when a response arrives (line 698). If a connection drops before all responses arrive, entries remain until the `PClient` is GC'd. Since `PClient` is captured by the connection thread which terminates on error, the `PClient` should become GC-eligible — but GC timing depends on heap pressure.
+
+#### 6. `subQ :: TQueue (ClientSub, ClientId)` — `Env/STM.hs:363`
+
+Unbounded `TQueue` for subscription changes. If the subscriber thread (`serverThread`) can't process changes fast enough, this queue grows without backpressure. With 22K clients subscribing/unsubscribing, sustained bursts could cause this queue to bloat.
+
+---
+
+### Ruled Out
+
+1. **PostgreSQL queue cache**: `useCache = False` — `queues`, `senders`, `links`, `notifiers` TMaps are empty.
+2. **`notifierLocks`**: Guarded by `useCache` (Postgres.hs:377,405) — not used with `useCache = False`.
+3. **Client structures**: 22K × ~3KB = ~66MB — negligible.
+4. **TBQueues**: Bounded (`tbqSize = 128`).
+5. **Thread management**: `forkClient` uses weak refs + `finally` blocks. `endThreads` cleared on disconnect.
+6. **Proxy `smpClients`/`smpSessions`**: Properly cleaned on disconnect/expiry.
+7. **`smpSubWorkers`**: Properly cleaned on worker completion; also cleared in `closeSMPClientAgent`.
+8. **`pendingEvents`**: Atomically swapped empty every `pendingENDInterval`.
+9. **Stats IORef counters**: Fixed number, bounded.
+10. **DB connection pool**: Bounded `TBQueue` with bracket-based return.
+
+---
+
+### Insufficient Data to Determine Root Cause
+
+Without measuring the actual sizes of these structures at runtime, we cannot determine which (if any) is the primary contributor. The following exact logging changes will identify the root cause.
+
+---
+
+### EXACT LOGS TO ADD
+
+Add a new periodic logging thread in `src/Simplex/Messaging/Server.hs`.
+
+Insert at `Server.hs:197` (after `prometheusMetricsThread_`):
+
+```haskell
+        <> memoryDiagThread_ cfg
+```
+
+Then define:
+
+```haskell
+    memoryDiagThread_ :: ServerConfig s -> [M s ()]
+    memoryDiagThread_ ServerConfig {prometheusInterval = Just _} =
+      [memoryDiagThread]
+    memoryDiagThread_ _ = []
+
+    memoryDiagThread :: M s ()
+    memoryDiagThread = do
+      labelMyThread "memoryDiag"
+      Env { ntfStore = NtfStore ntfMap
+          , server = srv@Server {subscribers, ntfSubscribers}
+          , proxyAgent = ProxyAgent {smpAgent = pa}
+          , msgStore_ = ms
+          } <- ask
+      let interval = 300_000_000 -- 5 minutes
+      liftIO $ forever $ do
+        threadDelay interval
+        -- GHC RTS stats
+        rts <- getRTSStats
+        let liveBytes = gcdetails_live_bytes $ gc rts
+            heapSize = gcdetails_mem_in_use_bytes $ gc rts
+            gcCount = gcs rts
+        -- Server structures
+        clientCount <- IM.size <$> getServerClients srv
+        -- SubscribedClients (queue and service subscribers for both SMP and NTF)
+        smpQSubs <- M.size <$> getSubscribedClients (queueSubscribers subscribers)
+        smpSSubs <- M.size <$> getSubscribedClients (serviceSubscribers subscribers)
+        ntfQSubs <- M.size <$> getSubscribedClients (queueSubscribers ntfSubscribers)
+        ntfSSubs <- M.size <$> getSubscribedClients (serviceSubscribers ntfSubscribers)
+        -- Pending events
+        smpPending <- IM.size <$> readTVarIO (pendingEvents subscribers)
+        ntfPending <- IM.size <$> readTVarIO (pendingEvents ntfSubscribers)
+        -- NtfStore
+        ntfStoreSize <- M.size <$> readTVarIO ntfMap
+        -- ProxyAgent maps
+        let SMPClientAgent {smpClients, smpSessions, activeServiceSubs, activeQueueSubs, pendingServiceSubs, pendingQueueSubs, smpSubWorkers} = pa
+        paClients <- M.size <$> readTVarIO smpClients
+        paSessions <- M.size <$> readTVarIO smpSessions
+        paActSvc <- M.size <$> readTVarIO activeServiceSubs
+        paActQ <- M.size <$> readTVarIO activeQueueSubs
+        paPndSvc <- M.size <$> readTVarIO pendingServiceSubs
+        paPndQ <- M.size <$> readTVarIO pendingQueueSubs
+        paWorkers <- M.size <$> readTVarIO smpSubWorkers
+        -- Loaded queue counts
+        lc <- loadedQueueCounts $ fromMsgStore ms
+        -- Log everything
+        logInfo $
+          "MEMORY "
+            <> "rts_live=" <> tshow liveBytes
+            <> " rts_heap=" <> tshow heapSize
+            <> " rts_gc=" <> tshow gcCount
+            <> " clients=" <> tshow clientCount
+            <> " smpQSubs=" <> tshow smpQSubs
+            <> " smpSSubs=" <> tshow smpSSubs
+            <> " ntfQSubs=" <> tshow ntfQSubs
+            <> " ntfSSubs=" <> tshow ntfSSubs
+            <> " smpPending=" <> tshow smpPending
+            <> " ntfPending=" <> tshow ntfPending
+            <> " ntfStore=" <> tshow ntfStoreSize
+            <> " paClients=" <> tshow paClients
+            <> " paSessions=" <> tshow paSessions
+            <> " paActSvc=" <> tshow paActSvc
+            <> " paActQ=" <> tshow paActQ
+            <> " paPndSvc=" <> tshow paPndSvc
+            <> " paPndQ=" <> tshow paPndQ
+            <> " paWorkers=" <> tshow paWorkers
+            <> " loadedQ=" <> tshow (loadedQueueCount lc)
+            <> " loadedNtf=" <> tshow (loadedNotifierCount lc)
+            <> " ntfLocks=" <> tshow (notifierLockCount lc)
+```
+
+Note: `smpSubs.subsCount` (queueSubscribers size) and `smpSubs.subServicesCount` (serviceSubscribers size) are **already logged** in Prometheus (lines 475-496). The log above adds all other candidate structures plus GHC RTS memory stats.
+
+This produces a single log line every 5 minutes:
+
+```
+[INFO] MEMORY rts_live=10737418240 rts_heap=23488102400 rts_gc=4521 clients=21927 smpQSubs=1847233 smpSSubs=42 ntfQSubs=982112 ntfSSubs=31 smpPending=0 ntfPending=0 ntfStore=512844 paClients=12 paSessions=12 paActSvc=0 paActQ=0 paPndSvc=0 paPndQ=0 paWorkers=3 loadedQ=0 loadedNtf=0 ntfLocks=0
+```
+
+### What Each Metric Tells Us
+
+| Metric | What it reveals | If growing = suspect |
+|--------|----------------|---------------------|
+| `rts_live` | Actual live data after last major GC | Baseline — everything else should add up to this |
+| `rts_heap` | Total heap (should be ~2× rts_live) | If >> 2× live, fragmentation issue |
+| `clients` | Connected client count | Known: ~22K |
+| `smpQSubs` | SubscribedClients map size (queue subs) | If >> clients × avg_subs, entries not cleaned |
+| `smpSSubs` | SubscribedClients map size (service subs) | Should be small |
+| `ntfQSubs` | NTF SubscribedClients map (queue subs) | Same concern as smpQSubs |
+| `ntfSSubs` | NTF SubscribedClients map (service subs) | Should be small |
+| `smpPending` / `ntfPending` | Pending END/DELD events per client | If large, subscriber thread lagging |
+| `ntfStore` | NotifierId count in NtfStore | If growing monotonically, zombie entries |
+| `paClients` | Proxy connections to other servers | Should be <= unique dest servers |
+| `paSessions` | Active proxy sessions | Should match paClients |
+| `paActSvc` / `paActQ` | Proxy active subscriptions | If growing, entries never removed |
+| `paPndSvc` / `paPndQ` | Proxy pending subscriptions | If growing, resubscription stuck |
+| `paWorkers` | Active reconnect workers | If growing, workers stuck in retry |
+| `loadedQ` | Cached queues in store (0 with useCache=False) | Should be 0 |
+| `ntfLocks` | Notifier locks in store | Should be 0 with useCache=False |
+
+### Interpretation Guide
+
+**If `smpQSubs` is in the millions**: SubscribedClients is the primary leak. Entries accumulate for every queue ever subscribed to.
+
+**If `ntfStore` grows monotonically**: Zombie notification entries (empty lists after expiration). Fix: `deleteExpiredNtfs` should remove entries with empty lists.
+
+**If `paActSvc` + `paActQ` grow**: Proxy agent subscription maps are the leak. Fix: add cleanup when no active/pending subs exist for a server.
+
+**If `rts_live` is much smaller than `rts_heap`**: GHC heap fragmentation. Fix: tune `-F` flag (GC trigger factor) or use `-c` (compacting GC).
+
+**If `rts_live` ~ 10-12GB**: The live data is genuinely large. Look at which metric is the largest contributor.
+
+**If nothing above is large but `rts_live` is large**: The leak is in a structure not measured here — likely TLS connection buffers, ByteString retention from Postgres queries, or GHC runtime overhead. Next step would be heap profiling with `-hT`.

simplexmq.cabal

@@ -432,6 +432,36 @@ executable smp-server
     , text
   default-language: Haskell2010
 
+executable smp-server-bench
+  if flag(client_library)
+      buildable: False
+  if flag(server_postgres)
+    cpp-options: -DdbServerPostgres
+  main-is: Main.hs
+  other-modules:
+      ClientSim
+      Report
+  hs-source-dirs:
+      bench
+  default-extensions:
+      StrictData
+  ghc-options: -O2 -threaded -rtsopts
+  build-depends:
+      base
+    , async
+    , bytestring
+    , containers
+    , crypton
+    , mtl
+    , network
+    , simple-logger
+    , simplexmq
+    , stm
+    , text
+    , time
+    , unliftio
+  default-language: Haskell2010
+
 executable xftp
   if flag(client_library)
       buildable: False

src/Simplex/Messaging/Server.hs

@@ -91,7 +91,7 @@ import qualified Data.X509 as X
 import qualified Data.X509.Validation as XV
 import GHC.Conc.Signal
 import GHC.IORef (atomicSwapIORef)
-import GHC.Stats (getRTSStats)
+import GHC.Stats (RTSStats (..), GCDetails (..), getRTSStats)
 import GHC.TypeLits (KnownNat)
 import Network.Socket (ServiceName, Socket, socketToHandle)
 import qualified Network.TLS as TLS
@@ -198,6 +198,7 @@ smpServer started cfg@ServerConfig {transports, transportConfig = tCfg, startOpt
             <> serverStatsThread_ cfg
             <> prometheusMetricsThread_ cfg
             <> controlPortThread_ cfg
+            <> [memoryDiagThread]
     )
     `finally` stopServer s
   where
@@ -719,6 +720,75 @@ smpServer started cfg@ServerConfig {transports, transportConfig = tCfg, startOpt
                 Nothing -> acc
                 Just (_, ts) -> (cnt + 1, updateTimeBuckets ts ts' times)
 
+    memoryDiagThread :: M s ()
+    memoryDiagThread = do
+      labelMyThread "memoryDiag"
+      Env
+        { ntfStore = NtfStore ntfMap,
+          server = srv@Server {subscribers, ntfSubscribers},
+          proxyAgent = ProxyAgent {smpAgent = pa},
+          msgStore_ = ms
+        } <- ask
+      let SMPClientAgent {smpClients, smpSessions, activeServiceSubs, activeQueueSubs, pendingServiceSubs, pendingQueueSubs, smpSubWorkers} = pa
+      liftIO $ forever $ do
+        threadDelay 300_000_000 -- 5 minutes
+        rts <- getRTSStats
+        let GCDetails {gcdetails_live_bytes, gcdetails_mem_in_use_bytes, gcdetails_large_objects_bytes, gcdetails_compact_bytes, gcdetails_block_fragmentation_bytes} = gc rts
+        clientCount <- IM.size <$> getServerClients srv
+        smpQSubs <- M.size <$> getSubscribedClients (queueSubscribers subscribers)
+        smpSSubs <- M.size <$> getSubscribedClients (serviceSubscribers subscribers)
+        ntfQSubs <- M.size <$> getSubscribedClients (queueSubscribers ntfSubscribers)
+        ntfSSubs <- M.size <$> getSubscribedClients (serviceSubscribers ntfSubscribers)
+        smpPending <- IM.size <$> readTVarIO (pendingEvents subscribers)
+        ntfPending <- IM.size <$> readTVarIO (pendingEvents ntfSubscribers)
+        ntfStoreSize <- M.size <$> readTVarIO ntfMap
+        paClients' <- M.size <$> readTVarIO smpClients
+        paSessions' <- M.size <$> readTVarIO smpSessions
+        paActSvc <- M.size <$> readTVarIO activeServiceSubs
+        paActQ <- M.size <$> readTVarIO activeQueueSubs
+        paPndSvc <- M.size <$> readTVarIO pendingServiceSubs
+        paPndQ <- M.size <$> readTVarIO pendingQueueSubs
+        paWorkers <- M.size <$> readTVarIO smpSubWorkers
+        lc <- loadedQueueCounts $ fromMsgStore ms
+        -- per-client metrics: total subscriptions and queue fill
+        clients <- getServerClients srv
+        let clientsList = IM.elems clients
+        totalSubs <- sum <$> mapM (\Client {subscriptions} -> M.size <$> readTVarIO subscriptions) clientsList
+        totalSndQ <- sum <$> mapM (\Client {sndQ} -> fromIntegral <$> atomically (lengthTBQueue sndQ)) clientsList
+        totalMsgQ <- sum <$> mapM (\Client {msgQ} -> fromIntegral <$> atomically (lengthTBQueue msgQ)) clientsList
+        totalEndThreads <- sum <$> mapM (\Client {endThreads} -> IM.size <$> readTVarIO endThreads) clientsList
+        logInfo $
+          "MEMORY"
+            <> " rts_live=" <> tshow gcdetails_live_bytes
+            <> " rts_heap=" <> tshow gcdetails_mem_in_use_bytes
+            <> " rts_max_live=" <> tshow (max_live_bytes rts)
+            <> " rts_large=" <> tshow gcdetails_large_objects_bytes
+            <> " rts_compact=" <> tshow gcdetails_compact_bytes
+            <> " rts_frag=" <> tshow gcdetails_block_fragmentation_bytes
+            <> " rts_gc=" <> tshow (gcs rts)
+            <> " clients=" <> tshow clientCount
+            <> " clientSubs=" <> tshow totalSubs
+            <> " clientSndQ=" <> tshow (totalSndQ :: Int)
+            <> " clientMsgQ=" <> tshow (totalMsgQ :: Int)
+            <> " clientThreads=" <> tshow totalEndThreads
+            <> " smpQSubs=" <> tshow smpQSubs
+            <> " smpSSubs=" <> tshow smpSSubs
+            <> " ntfQSubs=" <> tshow ntfQSubs
+            <> " ntfSSubs=" <> tshow ntfSSubs
+            <> " smpPending=" <> tshow smpPending
+            <> " ntfPending=" <> tshow ntfPending
+            <> " ntfStore=" <> tshow ntfStoreSize
+            <> " paClients=" <> tshow paClients'
+            <> " paSessions=" <> tshow paSessions'
+            <> " paActSvc=" <> tshow paActSvc
+            <> " paActQ=" <> tshow paActQ
+            <> " paPndSvc=" <> tshow paPndSvc
+            <> " paPndQ=" <> tshow paPndQ
+            <> " paWorkers=" <> tshow paWorkers
+            <> " loadedQ=" <> tshow (loadedQueueCount lc)
+            <> " loadedNtf=" <> tshow (loadedNotifierCount lc)
+            <> " ntfLocks=" <> tshow (notifierLockCount lc)
+
     runClient :: Transport c => X.CertificateChain -> C.APrivateSignKey -> TProxy c 'TServer -> c 'TServer -> M s ()
     runClient srvCert srvSignKey tp h = do
       ms <- asks msgStore