Test for benchmarks in multiple threads is not properly written

[Gustavo Narea]

Which causes that *sometimes*, we get failures like:

FAIL: test_running_readonly_benchmark_in_two_threads (tests.test_adapters_benchmark.TestAdapterBenchmark)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/gustavo/Software/Efous/Grupos/trunk/tests/test_adapters_benchmark.py", line 137, in test_running_readonly_benchmark_in_two_threads
    eq_(adapter.resets - 1, iterations * threads)
  File "/home/gustavo/System/pyenvs/grupos/lib/python2.6/site-packages/nose-0.11.3-py2.6.egg/nose/tools.py", line 31, in eq_
    assert a == b, msg or "%r != %r" % (a, b)
AssertionError: 9 != 10
    """Shorthand for 'assert a == b, "%r != %r" % (a, b)
    """
>> assert 9 == 10, None or "%r != %r" % (9, 10)