Definition: Big data testing is the testing of systems that store, process or analyse very large volumes of data, often across many servers. It covers whether data is processed correctly and whether the system performs and scales at production-like volumes. Because full-scale data can take weeks or months to load, testers often need creative ways to simulate scale.
So what? Traditional test phases with long load and performance runs don't fit short agile iterations. Big data testing pushes testers to understand the system's internals, such as indexes and metadata, so they can test at scale without always using full-scale data.
Examples: Building a small test archive whose metadata behaves like a multi-terabyte system. Restoring a pre-populated installation from backup so performance tests can start immediately.
