This paper presents industrial experiences on improving the fuzzing framework of SAP HANA to provide better configurability for continuous unit-level fuzzing. The fuzzing framework for unit-level testing should provide high configurability for the users to find and use the best configuration for each unit fuzzing driver beause each target component may have unique characteristics. To improve fuzzer configurability for unit-level testing, we provide new mutation scheduling strategies for effective uses of customized mutation operators, and new seed corpus selection strategies to configure a fuzzing campaign to check on changed code. The empirical results show that these extensions give better chances to the user to improve the effectiveness and the efficiency of the fuzzing.