Spark, Spark Streaming and Spark SQL unit testing strategies
Features
- ClockWrapper for efficient clock management in Spark Streaming jobs.
- Base traits for testing Spark, Spark Streaming and Spark SQL to eliminate boilerplate code.
- Sample applications to show how to make your code testable.
- All tests can be run or debugged directly from IDE, or using SBT.
- All test fixtures are prepared as in-memory data structures.
- SBT is configured to avoid problems with multiple Spark contexts in the same JVM SPARK-2243.
- Logger is configured, uncomment specific loggers in logback-test.xml if needed.