add: use SO_REUSEPORT on platform supporting it

Author: mkleczek

CHANGELOG.md

@@ -13,12 +13,17 @@ All notable changes to this project will be documented in this file. From versio
 - Optimize requests with `Prefer: count=exact` that do not use ranges or `db-max-rows` by @laurenceisla in #3957
   + Removed unnecessary double count when building the `Content-Range`.
 - Add config `client_error_verbosity` to customize error verbosity by @taimoorzaeem in #4088, #3980, #3824
+- Use SO_REUSEPORT on platforms supporting it by @mkleczek in #4703 #4694
 
 ### Changed
 
 - Log error when `db-schemas` config contains schema `pg_catalog` or `information_schema` by @taimoorzaeem in #4359
   + Now fails at startup. Prior to this, it failed with `PGRST205` on requests related to these schemas.
 
+### Fixed
+
+- Shutdown should wait for in flight requests by @mkleczek in #4702
+
 ## [14.6] - 2026-03-06
 
 ### Fixed

nix/overlays/haskell-packages.nix

@@ -78,6 +78,53 @@ let
       hasql-pool = lib.dontCheck prev.hasql-pool_1_0_1;
       hasql-transaction = lib.dontCheck prev.hasql-transaction_1_1_0_1;
       postgresql-binary = lib.dontCheck (lib.doJailbreak prev.postgresql-binary_0_13_1_3);
+
+      http2 =
+        prev.callHackageDirect
+          {
+            pkg = "http2";
+            ver = "5.4.0";
+            sha256 = "sha256-PeEWVd61bQ8G7LvfLeXklzXqNJFaAjE2ecRMWJZESPE=";
+          }
+          { };
+
+      http-semantics =
+        prev.callHackageDirect
+          {
+            pkg = "http-semantics";
+            ver = "0.4.0";
+            sha256 = "sha256-rh0z51EKvsu5rQd5n2z3fSRjjEObouNZSBPO9NFYOF0=";
+          }
+          { };
+
+      network-run =
+        prev.callHackageDirect
+          {
+            pkg = "network-run";
+            ver = "0.5.0";
+            sha256 = "sha256-vbXh+CzxDsGApjqHxCYf/ijpZtUCApFbkcF5gyN0THU=";
+          }
+          { };
+
+      time-manager =
+        prev.callHackageDirect
+          {
+            pkg = "time-manager";
+            ver = "0.2.4";
+            sha256 = "sha256-sAt/331YLQ2IU3z90aKYSq1nxoazv87irsuJp7ZG3pw=";
+          }
+          { };
+
+      warp =
+        lib.dontCheck (prev.callCabal2nixWithOptions "warp"
+          (super.fetchFromGitHub {
+            owner = "mkleczek";
+            repo = "wai";
+            rev = "e9c9e784dee1b54461b94089175a969eb647d262";
+            #sha256 = "sha256-AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=";
+            sha256 = "sha256-q8vgvsNeKM/fHp3H6+W/tR4HicioqeaU9Eo3rb13bLo=";
+          }) "--subpath=warp"
+          { });
     };
 in
 {

src/PostgREST/Admin.hs

@@ -22,20 +22,20 @@ import qualified PostgREST.AppState as AppState
 import qualified Network.Socket as NS
 import           Protolude
 
-runAdmin :: AppState -> Maybe NS.Socket -> NS.Socket -> Warp.Settings -> IO ()
-runAdmin appState maybeAdminSocket socketREST settings = do
+runAdmin :: AppState -> Maybe NS.Socket -> IO (Maybe NS.Socket) -> Warp.Settings -> IO ()
+runAdmin appState maybeAdminSocket getSocketREST settings = do
   whenJust maybeAdminSocket $ \adminSocket -> do
     address <- resolveSocketToAddress adminSocket
     observer $ AdminStartObs address
     void . forkIO $ Warp.runSettingsSocket settings adminSocket adminApp
   where
-    adminApp = admin appState socketREST
+    adminApp = admin appState getSocketREST
     observer = AppState.getObserver appState
 
 -- | PostgREST admin application
-admin :: AppState.AppState -> NS.Socket -> Wai.Application
-admin appState socketREST req respond  = do
-  isMainAppReachable  <- isRight <$> reachMainApp socketREST
+admin :: AppState.AppState -> IO (Maybe NS.Socket) -> Wai.Application
+admin appState getSocketREST req respond = do
+  isMainAppReachable <- getSocketREST >>= maybe (pure False) (fmap isRight . reachMainApp)
   isLoaded <- AppState.isLoaded appState
   isPending <- AppState.isPending appState
 
@@ -44,8 +44,8 @@ admin appState socketREST req respond  = do
       respond $ Wai.responseLBS (if isMainAppReachable then HTTP.status200 else HTTP.status500) [] mempty
     ["ready"] ->
       let
-        status | not isMainAppReachable = HTTP.status500
-               | isPending              = HTTP.status503
+        status | isPending              = HTTP.status503
+               | not isMainAppReachable = HTTP.status500
                | isLoaded               = HTTP.status200
                | otherwise              = HTTP.status500
       in

src/PostgREST/App.hs

@@ -22,6 +22,9 @@ import GHC.IO.Exception (IOErrorType (..))
 import System.IO.Error  (ioeGetErrorType)
 
 import Control.Monad.Except     (liftEither)
+import Control.Monad.Extra      (whenJust)
+import Data.IORef               (atomicWriteIORef, newIORef,
+                                 readIORef)
 import Data.Either.Combinators  (mapLeft, whenLeft)
 import Data.Maybe               (fromJust)
 import Data.String              (IsString (..))
@@ -62,8 +65,8 @@ import PostgREST.Version              (docsVersion, prettyVersion)
 
 import qualified Data.ByteString.Char8  as BS
 import qualified Data.List              as L
-import           Data.Streaming.Network (bindPortTCP,
-                                         bindRandomPortTCP)
+import           Data.Streaming.Network (HostPreference,
+                                         bindPortGenEx)
 import qualified Data.Text              as T
 import qualified Network.HTTP.Types     as HTTP
 import qualified Network.Socket         as NS
@@ -76,21 +79,36 @@ run :: AppState -> IO ()
 run appState = do
   conf@AppConfig{..} <- AppState.getConfig appState
 
-  AppState.schemaCacheLoader appState -- Loads the initial SchemaCache
-  (mainSocket, adminSocket) <- initSockets conf
+  mainSocketRef <- newIORef Nothing
+  adminSocket <- initAdminServerSocket conf
 
-  Unix.installSignalHandlers (AppState.getMainThreadId appState) (AppState.schemaCacheLoader appState) (AppState.readInDbConfig False appState)
+  let closeSockets = do
+        whenJust adminSocket NS.close
+        readIORef mainSocketRef >>= foldMap NS.close
+  Unix.installSignalHandlers closeSockets (AppState.schemaCacheLoader appState) (AppState.readInDbConfig False appState)
+
+  Admin.runAdmin appState adminSocket (readIORef mainSocketRef) (serverSettings conf)
 
   Listener.runListener appState
 
-  Admin.runAdmin appState adminSocket mainSocket (serverSettings conf)
+  -- Kick off and wait for the initial SchemaCache load before creating the
+  -- main API socket.
+  AppState.schemaCacheLoader appState
+  AppState.waitForSchemaCacheLoaded appState
+
+  mainSocket <- initServerSocket conf
+  atomicWriteIORef mainSocketRef $ Just mainSocket
 
   let app = postgrest configLogLevel appState (AppState.schemaCacheLoader appState)
 
   do
     address <- resolveSocketToAddress mainSocket
     observer $ AppServerAddressObs address
 
+  -- Hardcoding maximum graceful shutdown timeout (arbitrary set to 5 seconds)
+  -- This is unfortunate but necessary becase graceful shutdowns don't work with HTTP keep-alive
+  -- causing Warp to handle requests on already opened connections even if the listen socket is closed
+  -- See: https://github.com/yesodweb/wai/issues/853
   Warp.runSettingsSocket (serverSettings conf & setOnException onWarpException) mainSocket app
   where
     observer = AppState.getObserver appState
@@ -237,39 +255,27 @@ addRetryHint delay response = do
 isServiceUnavailable :: Wai.Response -> Bool
 isServiceUnavailable response = Wai.responseStatus response == HTTP.status503
 
-type AppSockets = (NS.Socket, Maybe NS.Socket)
-
-initSockets :: AppConfig -> IO AppSockets
-initSockets AppConfig{..} = do
-  let
-    cfg'usp = configServerUnixSocket
-    cfg'uspm = configServerUnixSocketMode
-    cfg'host = configServerHost
-    cfg'port = configServerPort
-    cfg'adminHost = configAdminServerHost
-    cfg'adminPort = configAdminServerPort
-
-  sock <- case cfg'usp of
-    -- I'm not using `streaming-commons`' bindPath function here because it's not defined for Windows,
-    -- but we need to have runtime error if we try to use it in Windows, not compile time error
-    Just path -> createAndBindDomainSocket path cfg'uspm
-    Nothing -> do
-      (_, sock) <-
-        if cfg'port /= 0
-          then do
-            sock <- bindPortTCP cfg'port (fromString $ T.unpack cfg'host)
-            pure (cfg'port, sock)
-          else do
-            -- explicitly bind to a random port, returning bound port number
-            (num, sock) <- bindRandomPortTCP (fromString $ T.unpack cfg'host)
-            pure (num, sock)
-      pure sock
-
-  adminSock <- case cfg'adminPort of
-    Just adminPort -> do
-      adminSock <- bindPortTCP adminPort (fromString $ T.unpack cfg'adminHost)
-      pure $ Just adminSock
-    Nothing -> pure Nothing
-
-  pure (sock, adminSock)
-
+initServerSocket :: AppConfig -> IO NS.Socket
+initServerSocket AppConfig{..} = case configServerUnixSocket of
+  -- I'm not using `streaming-commons`' bindPath function here because it's not defined for Windows,
+  -- but we need to have runtime error if we try to use it in Windows, not compile time error
+  Just path -> createAndBindDomainSocket path configServerUnixSocketMode
+  Nothing ->
+    bindPortTCPWithReusePort configServerPort (fromString $ T.unpack configServerHost)
+
+initAdminServerSocket :: AppConfig -> IO (Maybe NS.Socket)
+initAdminServerSocket AppConfig{..} =
+  traverse (`bindPortTCPWithReusePort` adminHost) configAdminServerPort
+  where
+    adminHost = fromString $ T.unpack configAdminServerHost
+
+bindPortTCPWithReusePort :: Int -> HostPreference -> IO NS.Socket
+bindPortTCPWithReusePort port hostPreference = do
+  -- Some unix variants can expose ReusePort but reject it at runtime.
+  -- Fall back to binding without ReusePort when that happens.
+  try (bindPortGenEx reusePortOpts NS.Stream port hostPreference) :: IO (Either SomeException NS.Socket)
+  >>= either (const $ bindPortGenEx [] NS.Stream port hostPreference) pure
+  >>= listenSocket
+  where
+    reusePortOpts = [(NS.ReusePort, 1)]
+    listenSocket sock = NS.listen sock (max 2048 NS.maxListenQueue) $> sock

src/PostgREST/AppState.hs

@@ -25,6 +25,7 @@ module PostgREST.AppState
   , getObserver
   , isLoaded
   , isPending
+  , waitForSchemaCacheLoaded
   ) where
 
 import qualified Data.ByteString.Char8      as BS
@@ -384,6 +385,9 @@ markSchemaCacheLoaded = void . (`tryPutMVar` ()) . getSCStatusMVar . stateSCache
 isSchemaCacheLoaded :: AppState -> IO Bool
 isSchemaCacheLoaded = fmap not . isEmptyMVar . getSCStatusMVar . stateSCacheStatus
 
+waitForSchemaCacheLoaded :: AppState -> IO ()
+waitForSchemaCacheLoaded = void . readMVar . getSCStatusMVar . stateSCacheStatus
+
 -- | Reads the in-db config and reads the config file again
 -- | We don't retry reading the in-db config after it fails immediately, because it could have user errors. We just report the error and continue.
 readInDbConfig :: Bool -> AppState -> IO ()

src/PostgREST/Unix.hs

@@ -18,10 +18,9 @@ import           System.Directory (removeFile)
 import           System.IO.Error  (isDoesNotExistError)
 
 -- | Set signal handlers, only for systems with signals
-installSignalHandlers :: ThreadId -> IO () -> IO () -> IO ()
+installSignalHandlers :: IO () -> IO () -> IO () -> IO ()
 #ifndef mingw32_HOST_OS
-installSignalHandlers tid usr1 usr2 = do
-  let interrupt = throwTo tid UserInterrupt
+installSignalHandlers interrupt usr1 usr2 = do
   install Signals.sigINT interrupt
   install Signals.sigTERM interrupt
   install Signals.sigUSR1 usr1

test/io/postgrest.py

@@ -88,7 +88,7 @@ def run(
     admin_port=None,
     host=None,
     wait_for_readiness=True,
-    wait_max_seconds=1,
+    wait_max_seconds=3,
     no_pool_connection_available=False,
     no_startup_stdout=True,
 ):
@@ -188,6 +188,7 @@ def wait_until_exit(postgrest, timeout=1):
 def wait_until_status_code(url, max_seconds, status_code):
     "Wait for the given HTTP endpoint to return a status code"
     session = requests_unixsocket.Session()
+    response = None
 
     for _ in range(max_seconds * 10):
         try: