
- Use go testing infrastructure, such that logs are printed only in case of failure. - Split the test into multiple sub-round and sub-round-testcases such that tree of execution reflects the logic. - Use 'development' loggers that are formatting the output to be human readable.