UltraFuzz: Towards Resource-Saving in Distributed Fuzzing

Recent research has sought to improve fuzzing performance via parallel computing. However, researchers focus on improving efficiency while ignoring the increasing cost of testing resources. Parallel in distributed environment amplifies resource-wasting problem caused by random nature fuzzing. In mode, owing lack an appropriate task dispatching scheme and timely status synchronization among different instances, conflicts workload imbalance occur, making severe. this paper, we design UltraFuzz, a fuzzer for resource-saving Based centralized dynamic scheduling, UltraFuzz can dispatch tasks schedule power globally reasonably avoid resource-wasting. Besides, elastically allocate computing seed evaluation, thereby avoiding potential bottleneck evaluation that blocks process. was evaluated using real-world programs, results show with same resource, outperforms state-of-the-art tools, such as AFL, AFL-P, PAFL, EnFuzz. Most importantly, experiment reveals certain seem counter-intuitive, namely achieve “super-linear acceleration” when compared single-core We conduct additional experiments reveal deep reasons behind phenomenon dig into inherent advantages over serial fuzzing, including global optimization energy scheduling escape local optimal seed. Additionally, 24 vulnerabilities were discovered UltraFuzz.