Integration and Interoperability Versus Functional Testing
The goal of integration and interoperability testing is to discover the incompatibilities described previously. Although it is not clear how much of a given application must be exercised to ensure complete integration coverage, the focus of this testing is to not duplicate any prior functional testing. Ideally, QA knows which functions result in communications external to the application and focuses testing on those functions. Without this detailed knowledge, QA performs basic integration and interoperability tests by ensuring that the fundamental functions of an application execute properly.
All defects are entered into Sun's BugTraq+ tool. Weekly "bug court" reviews are held to assess the assigned defects priority and severity level, and to determine if a given defect must be fixed prior to the next release or by General Availability (GA). Any necessary corrections are integrated into a new build of the software stack for the next test cycle. After all planned tests are executed at least once on the current build, the integration and interoperability tests on the entire stack can begin. Tests are then run to verify each correction (defect correction verification) and appropriate regression testing is run, according to QA Engineering's discretion.
This cycle of test, bug fix, and build continues until the exit criteria have been met. the new build is turned over to the Software Stack Test Group by IPG Release Engineering.
Typical Solaris OE Train Build Schedule
FIGURE 2 shows a typical Solaris OE Train build schedule for Sun Cluster and software stack testing.
FIGURE 2 Typical Solaris OE Build Schedule