Hi Apache Commons Pool community, I hope this email finds you well. I'd like to take feedback on a performance enhancement for Commons Pool, and if my understanding is correct.
*Problem Statement* While analyzing pool performance under high load, I identified that detailed statistics collection (mean active time, mean idle time, mean borrow wait time) can become a bottleneck in high-throughput scenarios. These statistics, while valuable for monitoring, involve timing calculations and circular buffer operations that add overhead to every borrow/return operation. For applications that prioritize throughput over detailed monitoring, there is no way to disable this overhead, and it leads to useful threads blocked on the synchronized block. [image: Screenshot 2025-08-22 at 1.40.11 AM.png] Proposed SolutionImplement a new configuration option collectDetailedStatistics that allows users to disable detailed timing statistics collection. I'm happy to discuss any aspects of this proposal or make adjustments based on community feedback. Also, I apologize in advance if I've missed any basic understanding of updateStatsBorrow/Return, and this change would impact any functional behaviour of the Pool. Any guidance or corrections would be greatly appreciated. And if this looks fine, I'll go ahead and raise the PR. Thanks.
