Jason Huynh created GEODE-486:
---------------------------------

             Summary: ConcurrentIndexUpdateWithWLDUnitTest needs to be 
rewritten/cleaned up
                 Key: GEODE-486
                 URL: https://issues.apache.org/jira/browse/GEODE-486
             Project: Geode
          Issue Type: Improvement
          Components: tests
            Reporter: Jason Huynh


This test has misleading variables such as parameters being passed in as  
30*000 (which obviously is 0 but this is supposed to be a parameter for join 
wait time, which means it waits infinitely and not that it doesn't wait at all)

This test along with any other test that uses PRQueryDUnitHelper should be 
written to not use the randomOps runnable.  Due to the nature of the test, the 
test exhausts different scenarios based on randomness and can fail/pass 
intermittently and fail for possibly different reasons.

Perhaps we can write a test to test the various scenarios we are hoping to 
exhaust with these random operations.  Such as test multiple puts on same keys 
for a specific index and make sure the index is maintained correctly.  Another 
test could be test multiple puts on different keys etc.

We may be able to have an overarching test that does these random operations 
but I don't think we should have a lot of these it's almost like writing the 
same test multiple times.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to