The Panasonic Distributed Testing Framework enables automated testing across multiple software versions and network topologies.
Testing functionality and interoperability between many versions of software components can be daunting. Since the number of tests grows exponentially with each version, the challenge of creating adequate test plans may even discourage modularized software development. The Panasonic Distributed Testing Framework helps solve this problem by automating the task of generating and executing tests in a scalable way.
Tests can be generated through scripts that, when executed, would create tests that exercise functionality across a combination of different versions or implementations of a software component. For example, a test generator script could specify that for every combination of programs called “bin/client” and “bin/server,” allocate a Linux machine and deploy the programs, then run server, run client, and check the results. If there were three libraries in different languages and five releases of each library, the generator would create 225 tests that would cover every combination.
Tests require resources. A test that requires a Linux machine could be executed on an Amazon AWS Instance, a Docker container, or on a server running in your office. A cluster of AWS instances running Docker could grow and shrink with demand, hosting as many Docker containers as needed. Resources can also include people, so the system can define and schedule manual tests as well as automated tests.
Once tests have been generated, a test executor prioritizes them, and then allocates resources and creates the necessary networks to produce the test results. Tests may be prioritized based on a release schedule or even by availability of resources.