Regression and Functional Testing for the iSeries

Extract coherent subsetted data

Creating test databases that are cutdown from the full production database is a must for both test performance and disk capacity reasons. And often it is necessary to have many copies of the test database available, with repeated executions of long running jobs.

Of crucial importance in these test databases is that the data maintain its complete relational integrity across all tables and all records.

Within X-Data Test are a set of functions to help you subset your database in a focused manner while maintaining its integrity.

For example, you might first specify that for testing this project you want a full set of canceled orders. By applying a filter you can subset the order file.

The next step is to create all the necessary related tables and records in the database. For example, you will probably need a set of customer master records that connect to the orders. And a set of item master records that connect to the order details. Those customer and item files may have codes that connect to keys in other tables, and so on.

By using the data model repository in X-Analysis, and in particular its deep knowledge of foreign keys, the subsetting function is able to deduce all needed tables and extract all necessary records for maintaining full relational integrity.

Mask sensitive, private data

Many countries require by law that sensitize consumer data be concealed from casual viewers such as data testers. Additionally, many companies felt the need to conceal sensitive, proprietary or confidential data – this is particularly true when testing activities are being outsourced.

A set of functions in X-Data Test allows you to scramble sensitive data while maintaining both its relational integrity and a visual appearance that users are accustomed to. For example, phone numbers still look like phone numbers, person's names look like actual names.

Design Intelligent Test Plans and Test Cases

It's stating the obvious, but how can you test what you don't know? Long-standing applications lack the documentation that are needed to develop test plans and detailed test cases.

Because X-Data Test is built on the X-Analysis platform, you are able to perform deep navigation and extraction of your application's business rules, data models, and even data content. There has never been a better way to intelligently construct the test cases you need for both regression testing and functionality changes.

By using the X-Analysis Business Rule repository in English-like pseudocode, test case developers are able to view the full sweep of business rules in the existing application and construct test cases that deliver a high degree of regression coverage.

There are also functions in X-Data Test that assist you with identifying the maximum extent of data values in your database and whether you have implemented them in your test database.  This helps you feel assured that you are testing as much of the system as possible with your test data.

Through use of the intuitive GUI, test case developers can navigate business processes, business rules, the data model, screen views, and drill down into actual data content in the system. This intuitive fluidity of navigation enables high quality, rapid development of test cases.

Manage Libraries of Test Runs

Through a process of defining test nodes, checkpoints and test results, X-Data Test helps you define and track repeated, consistent tests of both your baseline/current version code, and your new version, including repeated runs reflecting the testing and refinement of your new code.

Run Batch Tests Or Play Interactive Scripts

Through the use of test harnesses, X-Data Test enables you to define, manage and track repeated executions of batch code to be tested, whether entire batch jobs, or segmented portions.

Consistent and repeated execution of interactive programs is a major challenge of a different sort in regression testing. X-Data Test utilizes a technique known as image based testing to capture keyboard and mouse events and play them back, on demand, as often as needed. Such playback requires coordination with test database states, data content and definition of test cases.

After playback, X-Data Test compares the results on the screen using up to four different comparison algorithms. Additionally execution of playback may be logically conditioned based on test results.

Compare Thousands Of Fields Instantly In Files Or Reports

To achieve thorough testing of legacy applications it is often necessary to process high volumes of transactions for maximal test coverage. Analyzing the results of database output and report output can involve many thousands of individual field comparisons.

X-Data Test provides you with a high level view of test results and comparisons enabling you to quickly home in on potential defects.

Important to this sort of analysis is the ability to filter out expected and unavoidable differences for such things as timestamps. X-Data Test gives you complete flexibility for filtering out expected discrepancies so you can focus your energy and time on meaningful indicators of defects.

Detailed Drilldown Into Error Exploration

X-Data Test provides navigational and drilldown functionality to help you quickly and assuredly find your way through result comparison analysis. High level comparison starts with record level differences and allows drilldown into field level differences. If you are journaling your files, X-Data Test provides an easy way to research related journal entries to identify potentially defective programs and jobs.