test_2_File_Concurrency fails intermittently #1306
Description
Originally reported by: Nathan Kinder (Bitbucket: nkinder, GitHub: nkinder)
I've noticed intermittent failures of the test_2_File_Concurrency test in test_session.py. I'm using a recent checkout from the default branch (CherryPy 3.2.5 plus whatever has been committed since) with Python 2.7.5.
The failures can be seen by simply running the test_session.py tests in a loop as follows:
for i in `seq 1 100`; do nosetests -s ./test_session.py; done
I'm seeing failures about 5% of the time. Here is an example of the failure output:
S..........
----------------------------------------------------------------------
Ran 11 tests in 10.387s
OK (SKIP=1)
S...F......
======================================================================
FAIL: test_2_File_Concurrency (cherrypy.test.test_session.SessionTest)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/nathank/Source/cherrypy/cherrypy/cherrypy/test/test_session.py", line 218, in test_2_File_Concurrency
self._test_Concurrency()
File "/home/nathank/Source/cherrypy/cherrypy/cherrypy/test/test_session.py", line 268, in _test_Concurrency
self.assertEqual(hitcount, expected)
AssertionError: 150 != 151
-------------------- >> begin captured logging << --------------------
cherrypy.error: INFO: [19/Mar/2014:14:41:18] ENGINE Stopped thread 'Session cleanup'.
cherrypy.access.49232208: INFO: 127.0.0.1 - - [19/Mar/2014:14:41:18] "GET /setsessiontype/file HTTP/1.1" 200 - "" ""
cherrypy.error: INFO: [19/Mar/2014:14:41:18] ENGINE Started monitor thread 'Session cleanup'.
cherrypy.access.49232208: INFO: 127.0.0.1 - - [19/Mar/2014:14:41:18] "GET / HTTP/1.1" 200 1 "" ""
cherrypy.access.49232208: INFO: 127.0.0.1 - - [19/Mar/2014:14:41:18] "GET / HTTP/1.1" 200 1 "" ""
...
cherrypy.access.49232208: INFO: 127.0.0.1 - - [19/Mar/2014:14:41:20] "GET / HTTP/1.1" 200 3 "" ""
cherrypy.access.49232208: INFO: 127.0.0.1 - - [19/Mar/2014:14:41:20] "GET / HTTP/1.1" 200 3 "" ""
--------------------- >> end captured logging << ---------------------
----------------------------------------------------------------------
Ran 11 tests in 10.406s
FAILED (SKIP=1, failures=1)
S..........
----------------------------------------------------------------------
Ran 11 tests in 10.505s
OK (SKIP=1)
This looks very similar to ticket #799, which was resolved long ago.