On Oct 8, 2013, at 7:36 AM, Chip Childers wrote:

> On Tue, Oct 08, 2013 at 07:59:24PM +0530, Prasanna Santhanam wrote:
>> On Tue, Oct 08, 2013 at 10:23:32AM -0400, Chip Childers wrote:
>>> 3) I'd love if we could get some consensus on what additional tests and /
>>> or changes to the test approach are needed.  Prasanna - as with Hugo and
>>> Chiradeep, you've been "randomly" selected to at least provide some input
>>> here.
>> 
>> I saw the thread earlier about a mysql db generated for performing an
>> integration test. If someone can point me to the spec/docs/readme on
>> how to run these presumably without the contrail device I'm happy to
>> take a look.

The integration tests are being automatically executed by maven.
(mvn -pl :cloud-plugin-network-contrail clean test).

They spawn an instance of mysql on a dynamically allocated port in order to 
ensure that the database contents are always initialized with the
same content and that these set of tests do not leave content in the database 
that could influence other tests...

> 
> Pedro, perhaps you can provide some guidance here.
> 
> Prasanna - it appears that most of the stuff is in
> plugins/network-elements/juniper-contrail/test/
> 
correct.

> There are mysql start and stop .sh scripts, and unit tests in there as
> well.
> 

Yes. the shell scripts start and stop a mysql instance. The scripts are invoked 
from JUnit static initializer (@BeforeClass).

> -chip

Reply via email to