diff --git a/CHANGELOG.md b/CHANGELOG.md
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,4 +1,9 @@
 
+Version 0.7.4
+-----
+
+* A more graceful termination of the time server in case of self-destruction by time-out.
+
 Version 0.7.3
 -----
 
diff --git a/Simulation/Aivika/Distributed/Optimistic/Internal/TimeServer.hs b/Simulation/Aivika/Distributed/Optimistic/Internal/TimeServer.hs
--- a/Simulation/Aivika/Distributed/Optimistic/Internal/TimeServer.hs
+++ b/Simulation/Aivika/Distributed/Optimistic/Internal/TimeServer.hs
@@ -105,6 +105,8 @@
                -- ^ whether the time server is in the initial mode
                tsTerminating :: IORef Bool,
                -- ^ whether the time server is in the terminating mode
+               tsTerminated :: IORef Bool,
+               -- ^ whether the server is terminated
                tsProcesses :: IORef (M.Map DP.ProcessId LogicalProcessInfo),
                -- ^ the information about logical processes
                tsProcessesInFind :: IORef (S.Set DP.ProcessId),
@@ -156,6 +158,7 @@
 newTimeServer n ps =
   do f  <- newIORef True
      ft <- newIORef False
+     fe <- newIORef False
      m  <- newIORef M.empty
      s  <- newIORef S.empty
      t0 <- newIORef Nothing
@@ -165,6 +168,7 @@
                          tsInitQuorum = n,
                          tsInInit = f,
                          tsTerminating = ft,
+                         tsTerminated = fe,
                          tsProcesses = m,
                          tsProcessesInFind = s,
                          tsGlobalTime = t0,
@@ -530,6 +534,11 @@
                 then do tryComputeTimeServerGlobalTime server
                         loop utc
                 else loop utc0
+         loop' utc0 =
+           C.finally
+           (loop utc0)
+           (liftIO $
+            atomicWriteIORef (tsTerminated server) True)
      case tsSimulationMonitoringAction env of
        Nothing  -> return ()
        Just act ->
@@ -537,7 +546,7 @@
             monitorId <-
               DP.spawnLocal $
               let loop =
-                    do f <- liftIO $ readIORef (tsTerminating server)
+                    do f <- liftIO $ readIORef (tsTerminated server)
                        unless f $
                          do x <- DP.expectTimeout (tsSimulationMonitoringTimeout ps)
                             case x of
@@ -547,7 +556,7 @@
               in C.catch loop (handleTimeServerException server)
             DP.spawnLocal $
               let loop =
-                    do f <- liftIO $ readIORef (tsTerminating server)
+                    do f <- liftIO $ readIORef (tsTerminated server)
                        unless f $
                          do liftIO $
                               threadDelay (tsSimulationMonitoringInterval ps)
@@ -555,7 +564,7 @@
                             loop
               in C.catch loop (handleTimeServerException server)
             return ()
-     C.catch (liftIO getCurrentTime >>= loop) (handleTimeServerException server) 
+     C.catch (liftIO getCurrentTime >>= loop') (handleTimeServerException server) 
 
 -- | Handle the process monitor notification.
 handleProcessMonitorNotification :: DP.ProcessMonitorNotification -> TimeServer -> DP.Process ()
diff --git a/aivika-distributed.cabal b/aivika-distributed.cabal
--- a/aivika-distributed.cabal
+++ b/aivika-distributed.cabal
@@ -1,5 +1,5 @@
 name:            aivika-distributed
-version:         0.7.3
+version:         0.7.4
 synopsis:        Parallel distributed discrete event simulation module for the Aivika library
 description:
     This package extends the aivika-transformers [1] package and allows running parallel distributed simulations.
@@ -25,7 +25,7 @@
     or a set of reports consisting of HTML pages with charts, histograms, links to CSV tables, summary statistics etc.
     Please consult the AivikaSoft [3] website for more details.
     .
-    Regarding the speed of simulation, the rough estimations are as follows. The distributed simulation module is slower up to
+    Regarding the speed of simulation, the rough estimation is as follows. The distributed simulation module is slower up to
     6-9 times in comparison with the sequential aivika [2] simulation library using the equivalent sequential models. 
     Note that you can run up to 7 parallel logical processes on a single 8-core processor computer and run the Time Server 
     process too. On a 36-core processor, you can launch up to 35 logical processes simultaneously.