[Dev] Thoughts on service monitor and functional test definition

[Frank Cohen]

I've been doing some thinking about how PushToTest Version 5 could  
operate a functional test as a service monitor. Here is an example  
TestScenario XML snippet defining a load test:

   <testusecase>
     <dimensions>

       <!-- Concurrent request levels for each TestNode -->
       <crlevels>
         <crlevel value="10"/>
         <crlevel value="22"/>
       </crlevels>

       <usecases>
         <usecase name="Test_Mail_Service">

             <test>
               <run name="emailsend"  
testclass="com.examples.FunctionMail"
                       method="testFunctionMail" langtype="java"/>
             </test>
         </usecase>

Now here is a snippet defining a service monitor that uses the same  
functional test.

<monitor repeat=”60000”>
   <until>
     <anyexception>
     <testtime exceeds=”2000”>
   </until>

   <testusecase>
     <dimensions>

       <usecases>
         <usecase name="Test_Mail_Service">

             <test>
               <run name="emailsend"  
testclass="com.examples.FunctionMail"
                       method="testFunctionMail" langtype="java"/>
             </test>
         </usecase>

The difference is the <monitor> element and its children. Seems  
pretty straight forward.

I also wondered about running a functional test across a variety of  
TestNodes. Here is a snippet of a functional test document:

<functionaltest repeat=”1”>

   <testusecase>
     <dimensions>

       <usecases>
         <usecase name="Test_Mail_Service">

             <test>
               <run name="emailsend"  
testclass="com.examples.FunctionMail"
                       method="testFunctionMail" langtype="java"/>
             </test>
         </usecase>

What do you think? I'm looking for comments, feedback, and improvements.

-Frank


--
Frank Cohen, PushToTest, http://www.PushToTest.com, phone 408 374 7426
TestMaker: The open-source SOA test automation tool