The insertion points of the acquire() and release() methods determine the entire execution of the code. For this reason, it is very important that you take the time to analyze what threads you want to use and how you want to synchronize them.

For example, we can change the insertion point of the release() method in the MyThreadClass class like so:

import threading
import time
import os
from threading import Thread
from random import randint

# Lock Definition
threadLock = threading.Lock()

class MyThreadClass (Thread):
   def __init__(self, name, duration):
      Thread.__init__(self)
      self.name = name
      self.duration = duration
   def run(self):
      #Acquire the Lock
      threadLock.acquire() 
      print ("---> " + self.name + 
             " running, belonging to process ID " 
             + str(os.getpid()) + "
")
      #Release the Lock in this new point
      threadLock.release()
      time.sleep(self.duration)
      print ("---> " + self.name + " over
")

In this case, the output changes quite significantly:

---> Thread#1 running, belonging to process ID 11228
---> Thread#2 running, belonging to process ID 11228
---> Thread#3 running, belonging to process ID 11228
---> Thread#4 running, belonging to process ID 11228
---> Thread#5 running, belonging to process ID 11228
---> Thread#6 running, belonging to process ID 11228
---> Thread#7 running, belonging to process ID 11228
---> Thread#8 running, belonging to process ID 11228
---> Thread#9 running, belonging to process ID 11228

---> Thread#2 over
---> Thread#4 over
---> Thread#6 over
---> Thread#5 over
---> Thread#1 over
---> Thread#3 over
---> Thread#9 over
---> Thread#7 over
---> Thread#8 over

End
--- 6.11468243598938 seconds ---

As you can see, only the thread creation happens in sequential mode. Once thread creation is complete, the new thread acquires the lock, while the previous one continues the computation in the background.