Running several cassandra-unit tests is not possible

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Run more than one unit test that uses cassandra-unit for starting Cassandra

What is the expected output? What do you see instead?
While I expect all tests correctly run, only the first will succeed because 
from the second the embedded Cassandra complains about a duplicate index. This 
seems to be related to cassandra-unit which doesn't provide a way to shutdown 
the embedded instance between tests. 

Original issue reported on code.google.com by a.gazzarini@gmail.com on 31 Jan 2014 at 2:21

[GoogleCodeExporter]

With r813 most JUnit tests run again. However, some (important) tests still 
fail:
* SPARQLServletPojoTest
* HttpDeleteServiceOnTripleStoreTest
* RemoveTest
* QuadsTest
* QuadsDeletionTest
I will look into those the next days ...

- Andreas

Original comment by andreas.josef.wagner on 3 Feb 2014 at 1:50

• Changed state: Started

[GoogleCodeExporter]

Wow, nice shot!
For HttpDeleteServiceOnTripleStoreTest, at least locally, tests are 
green if you run them separately, so that should be something with setup 
/ teardown methods.

Original comment by a.gazzarini@gmail.com on 3 Feb 2014 at 5:55

[GoogleCodeExporter]

Ok,
HttpDeleteServiceOnTripleStoreTest now is green...that was my stupid 
mistake...testset iterator was static so once traversed in the first 
test method its hasNext always returned false.

Now, I will split CRUDServletPojoTest in one test case for each REST 
service (like HttpDeleteServiceOnTripleStoreTest is for DELETE)

Original comment by a.gazzarini@gmail.com on 3 Feb 2014 at 8:30

[GoogleCodeExporter]

I added a dedicated test case for POST REST service 
(HttpPostServiceOnTripleStore).

- all (both positive and negative) tests are green.
- test coverage indicates that we should run tests with all supported 
formats but since there's a todo (in CRUDServlet.doPost) for supporting 
as much formats as possible I think we take care about that later (I 
will open a separate issue / enhancement)

Original comment by a.gazzarini@gmail.com on 3 Feb 2014 at 10:39

[GoogleCodeExporter]

Added a dedicated test case for GET REST service 
(HttpGetServiceOnTripleStore).

- all (both positive and negative) tests are green.
- test coverage indicates that there are some branches not covered 
(mainly one branch for supported format)
- CRUDServletPojoiTest has been commented out because incompatible 
changes. When the last test (PUT) will be created it could be removed.

Original comment by a.gazzarini@gmail.com on 3 Feb 2014 at 10:52

[GoogleCodeExporter]

Added a dedicated test case for PUT REST service 
(HttpPutServiceOnTripleStore) with r861

Original comment by a.gazzarini@gmail.com on 6 Feb 2014 at 6:59

[GoogleCodeExporter]

Fixed with r892 : we have 148 green tests  

Original comment by a.gazzarini@gmail.com on 14 Feb 2014 at 4:07

• Changed state: Fixed

[GoogleCodeExporter]

Very cool :) Thanks for your good work. However, it seems we still have some 
Exceptions in the test runs. For some reason, we don't see this in the test 
outcomes?! I don't know why ... 

Attached find the console output from the build#5 ... The exceptions don't seem 
to major problems - some NullPointers, ParserExceptions etc. But we still 
should fix them.

Kind regards
Andreas

Original comment by andreas.josef.wagner on 16 Feb 2014 at 1:23

Attachments:

• cumulusRDFv1.0.1_build5.log

[GoogleCodeExporter]

This is also a strange ERROR in the above log (not an Execption, though):

ERROR [edu.kit.aifb.cumulus.util.hector.CassandraHectorCounterFactory] counter: 
TRIPLE_COUNTER suffered an overflow! current counter value: -3

This should not happen in the tests ;)

Original comment by andreas.josef.wagner on 16 Feb 2014 at 1:26

[GoogleCodeExporter]

... reopening the issue ... 

- Andreas

Original comment by andreas.josef.wagner on 16 Feb 2014 at 1:28

• Changed state: Started

[GoogleCodeExporter]

Hi Andreas, 
NPEs are ok, those test cases want that specific behaviour :D You can see in 
GET / POST / PUT (CRUDServlet) tests...for each of them there are two test 
cases that 

- try to see what happens in case of CumulusException: as commented there I'm 
not emulating each scenario that could raise that problem...just injecting a 
Store in ServletContext that is not yet open. The exception is caught, prnted 
out and the server sends a 500 INTERNAL SERVER ERROR as expected. You are 
seeing the exception because the logger on the servlet. 

- try to see what happens in case of *Exception: the catch (Exception) is the 
last block of each get/put/post method so here the test is seeing that if a 
(example) RuntimeException is thrown, the the expected result must be a 500 
INTERNAL SERVER ERROR. You are seeing the exception because the logger on the 
servlet. 

I'm going to see the other exception you mentioned.

Original comment by a.gazzarini@gmail.com on 16 Feb 2014 at 2:02

[GoogleCodeExporter]

After checking, all exceptions that are printed out during the execution of 
tests are expected, so I'm closing this issue.

Original comment by a.gazzarini@gmail.com on 1 Mar 2014 at 9:50

• Changed state: Fixed