[Zope] ZEO troubles on RedHat EL4 Linux

[Tim Peters]

[Andreas Krasa]

So you showed a number of test errors here, all on line 121 of
ConnectionTests.py.  They all seem indentical to the first one:

> ERROR: checkMultipleAddresses
> (ZEO.tests.testConnection.MappingStorageConnectionTests)
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>  File "/usr/local/src/__zope__/Zope-2.7.7-final/lib/python/ZEO/tests/ConnectionTests.py", line 121, in tearDown
>    os.waitpid(pid, 0)
> OSError: [Errno 10] No child processes

That's peculiar for several reasons, and-- sorry --I have no theory to offer.

Because the failure is in tearDown(), the test body has _completed_. 
If there's no other failure reported for this test, that means the
test succeeded.  Instead it's dying while it's tearing the test
framework down.

Some of the ZEO tests create additional processes.  The place it's
dying is in this teardown loop:

    def tearDown(self):
        """Try to cause the tests to halt"""

        ...

            for pid in self._pids:
                os.waitpid(pid, 0)  # here

The test driver is simply waiting for the other process(es) it spawned
to exit.  If there was a failure to create the other process(es), then
I'd expect a test to die long before teardown.  But if it did create
the other process(es), then I have no theory for why Linux waitpid()
would claim there are no child processes.

Maybe this is relevant (it's the only vaguely similar report I recall
seeing across years, but I don't think it reached a conclusion):

    http://mail.zope.org/pipermail/zope3-users/2005-July/000794.html