Clear created attributes when tearing down tests

unittest will keep each test case around after completion pinning any
objects references by attributes on the test case. This can bloat
the memory used during a full test suite run to the point where
the test suite will require much more memory than running all of
the services does. On systems witout lots of memory, test suite
failures will occur as test cases cannot fork to execute programs.

Clearing out __dict__ of any attributes that don't start with a _
(since some are needed by unittest itself) reduces memory
significantly. This does require the super class tearDown to be
called last in some cases however.

Change-Id: I0e660b6c2a77c5613c5a523ba0a64c3d7f6dd128

1 files changed, 22 insertions, 0 deletions

diff --git a/nova/testing/README.rst b/nova/testing/README.rst
index 036f1c77d..67fa33d1d 100644
--- a/nova/testing/README.rst
+++ b/nova/testing/README.rst
@@ -33,6 +33,17 @@ Using Fakes
 
 TBD
 
+test.TestCase
+-------------
+The TestCase class from nova.test (generally imported as test) will
+automatically manage self.stubs using the stubout module and self.mox
+using the mox module during the setUp step. They will automatically
+verify and clean up during the tearDown step.
+
+If using test.TestCase, calling the super class setUp is required and
+calling the super class tearDown is required to be last if tearDown
+is overriden.
+
 Writing Functional Tests
 ------------------------
 
@@ -42,3 +53,14 @@ Writing Integration Tests
 -------------------------
 
 TBD
+
+Tests and assertRaises
+----------------------
+When asserting that a test should raise an exception, test against the
+most specific exception possible. An overly broad exception type (like
+Exception) can mask errors in the unit test itself.
+
+Example::
+
+    self.assertRaises(exception.InstanceNotFound, db.instance_get_by_uuid,
+                      elevated, instance_uuid)