Post-Release Support and Regression Testing

March 20 10:00 2012 by Nataliia Vasylyna Print This Article

For effective post-release software product support, defects reported by clients should be analyzed to correct the underlying software bugs. Diagnosis testing is generally used to recreate the defect scenario and to predict the bug.

Software maintenance usually comprises activities to adapt the software to various operational environments, or to make better the product in all possible ways.

Different software product updates through new releases can be an enlargement to these support activities. The software with a big history and multitudinous foregoing releases are generally referred to as legacy products.

In this case special way of testing named regression testing is normally used to assure that beforehand supported software functions are not negatively influenced by the updates.

Regression testing is closely associated with specific testing stages for legacy products and for main software support activities.

For new product releases, we often go through a full cycle of software testing, with the main attention on regression testing normally on the integration testing sub-stages where old and new software elements are integrated.

For small updates and software maintenance, regression testing may compose the major part of software testing.

Methods for regression testing are generally include:

An investigation of dissimilarities between the precede and the current version of the software to choose from existing test cases and to define what new test cases are required.
The new test cases pay the main attention to 2 aspects:

the interplay involving old and new parts, which is equal to integration testing, but with the main attention to specific type of interfaces and interplays
the newly developed or modified part, which is basically the same as testing of new systems but on a smaller scale