Step by Step CA TDM: Test Match

Discussion created by findi01 Employee on Aug 10, 2016

Purpose of Test Match is to be able to find test data that fits your the test cases.

This is a step by step approach on how to create your first Test Match. Once created, you will be able to use it in Agile Requirement Designer, or other test tools in order to match and/or book your test data.


This tutorial is for advanced users that already know TDM and how to create/publish data pools based on subsets/clone. If you’re a beginner, I recommend you to start with « Step by Step CA TDM: TDOD   ».


Start TDM :

  1. At login, connect to:
    • Data Source : the database where is located the data you need to match/book
    • Data Target : the database where you will store your Test Match results (it could be the same as the data source).
      • If Ms SQL Server, it should use a DSN-Less ODBC connection
  2. Create a Subset <SUBSET> based on your DataSource
    • Select all tables required to have all selection criteria of your test data
    • Save the subset rule in your repository
  3. Create a view <VIEW> or data cube from your subset
    • Right-click on the <SUBSET> rule you saved the step before, and select option « Create Cube view »
    • Select a cube from Data Source
    • Right-click on each table to add the required columns for your selection criteria
      • Don’t forget to add also one or several key columns that will be used to lock the record booked
    • Click « OK » until you get a window with SQL script to create your view
    • Execute the script and test your <VIEW>
  4. Create a <TEST MART> table
    • Click on menu « Register... »  and register your view <VUE> in the repositoy
    • Click on menu « Actions for registered object »
    • Select your <VIEW> and execute action « Fully constrained DDL »
    • Copy & paste the generated DDL script DDL into a SQL window connected to your Data Target
    • Replace table prefix « CUBE_ » by « TEST_MART_ »
    • Add following columns (that serve for test data booking)
      • test_name varchar(50)
      • test_user varchar(50)
      • test_locking varchar(2)
      • job_id numeric(10)
    • Exécute the script to create the <TEST MART> table
  5. Fill the <TEST MART> table
    • Create a Data Pool <DP GENERATOR> from your view <VIEW> and standard cloning rules
    • Publish the Data Pool to your <TEST MART> table
      • In publish window, change Table Location to target <TEST MART>
      • Force publish by ignoring the différences between source and target schema (this is the columns we added for locking)
    • Check all data has been published in table <TEST MART>
  6. Create the <TEST MATCH> data pool
    • Click on menu « Register... » and register the <TEST MART> table in the repository
    • Create a new Data Pool of type « Test Match »
    • In sheet « Tables and Columns », mettre
      • Profile Name: select Data Target connection
      • Match Data in Table: select your table <TEST MART>
      • Report match in column: test_name
      • Report user in column: test_user
      • Report locking in column: test_locking
      • Report Job ID in column: job_id
    • Save and edit your Data Pool again. In sheet « Key Columns »,put
      • Report Key columns: these are the columns used for locking and informing tester of the data booked for him. It must uniquely identify the data you selected.
      • Primary Key columns: These are the key columns that uniquely identify the data booked
      • Data Attribute columns: these are the columns that will be displayed as selection criteria for your test match
  7. Test your new <TEST MATCH> Data Pool
    • Open your data pool, and add a new line to it
    • At minimum fill these columns
      • Test name: a unique name
      • Test active: put « Y »
      • Test repeater: put 1 to book only 1 row
      • Test conditions 1 to 20: put at least one condition like « L3_RA_STATUS = ‘OK’ » (if you double click on a condition, a helper window will appear). If you double-click on this cell, a helper window appear to build efficient query.
      • Test user: put a username
      • Test locking: put « LK » for locking (data is booked only for you), « RS » for Read Shareable (test data could be shared with others, for example you only read it)
    • Right-Click the Data Pool and select « Test Match » option
      • check box « Clear existing match »
      • for « Runtime Parameters », select « Limit to a single test » and « Simulate Testmatch"
    • Choose « Immediate » and check Results reports
    • It should open an HTML result page like this one
    • If you query your <TEST MART> table, it should contain booking information for identified records like here (look at the only cells that aren’t yellow on the right side)
  8. Now, you can do your match through Agile Designer or TDoD query