Tests: Overview and tasks
Tests are the fundamental building blocks of integration
and SOA testing. A test sends an input message to an external service
or system in a protocol-specific way and validates the reply against
an expected result.
A test can be added to an existing operation by using the menu or the Test Factory toolbar. Several different methods can be used for creating tests: by using MEP, from templates, by pasting, and from recorded events.
Note: If the operation is not configured to point to
a specific transport, the Test from Template, Test
from MEP, and Tests from MEP are
unavailable from the New, Tests menu.
After a test is created and named, it is opened in The editing panel with a palette of available actions that are displayed above the test tree (see Changing the contents of a test).
Once added to a test, a test action is fully editable and configurable. To open an action in a test, simply double-click it. More options are available from the action’s menu, which is displayed when you right-click the action.
Note: When executed, a test’s steps are processed
from top to bottom.
- Test phases
A new, empty test always consists of three basic phases: Initialize, Test Steps and Tear-down. - Test views
Two views are available for a test, the Technical View
and the Business View 
. - Creating a single empty test
You can create a single empty test in Rational Integration Tester, - Creating a single test by using MEP
An operation’s Message Exchange Pattern (MEP) defines the pattern, schema, and binding for messages that are exchanged by the operation. You can use this information to create a test that adheres to the rules defined in the MEP. - Creating multiple tests by using MEP
An operation’s Message Exchange Pattern (MEP) defines the pattern, schema, and binding for messages that are exchanged by the operation. Use this information to create multiple tests with the help of the MEP wizard. - Creating a test from a template
After you create a test template, you can quickly create new tests from it. - Creating tests by pasting
You can create one or more tests by pasting one or more lines of text that are currently on the clipboard. New tests are named according to the pasted text. Each line of text on the clipboard creates a unique test, and the tests are created within the component or folder that is selected when pasting. - Creating tests from recorded events
Tests can be created by using events that are captured in the Recording Studio. - Setting the Test properties by the input and output tags
Use the Properties tab to define which tags are to be available on the input and output interface for the selected test. - Overriding the inherited integration for change management applications
Rational Integration Tester can be integrated with OSLC-compliant change management systems for raising defects within the context of a Rational Integration Tester test asset. - Changing the contents of a test
You can add or edit test actions, change the test action type, enable or disable test actions, and save messages as requirements. - Filtering header values
If you are using a stub to test more than one web service from a single URL, you can use header values to distinguish between the services. - Monitoring a log file during a test
If a log file is configured on any of the bound infrastructure components of the project, that file can be monitored during the execution of a test. - Example: Monitoring a remote log file
In this example, you add a log file to a database server. With the log file added, you can then monitor that file when you run a test. - The Tag Data Store
The Tag Data Store, which is accessed from the menu anywhere within a test, is the repository and access point for all of the data tags that are available to a test. - Message cases
Message cases are used in Message Switch actions. To add a message case to an existing message switch, right-click the message switch and select Add Message Case from the menu. - Pass paths
You can define action sequences to be carried out when a subscription-based action passes (that is, Subscribe, Receive Request, and Receive Reply). - Failure paths
To specify that a test take different steps when a specific action fails, you can create a failure path and define sequences of actions for it. For example, a Publish action can fail to publish a message, or a subscriber can fail to subscriber properly or an incoming message might not pass validation. - Running or scheduling the running of a test
You initiate a test execution in the Test Factory, but the test is run in the Test Lab perspective. After you start a test, the Test Lab perspective is displayed, from where you monitor the test’s execution. A test can be run either in the current environment or in any other environment. - Creating performance tests
A performance test includes one or more instances of a Rational Integration Tester test plus a number of probes that measure the load on external systems (for example, messaging servers). The following describes how to create a performance test.
Related concepts: