AI RESEARCH

WAter: A Workload-Adaptive Knob Tuning System based on Workload Compression

arXiv CS.AI

ArXi:2603.28809v1 Announce Type: cross Selecting appropriate values for the configurable parameters of Database Management Systems (DBMS) to improve performance is a significant challenge. Recent machine learning (ML)-based tuning systems have shown strong potential, but their practical adoption is often limited by the high tuning cost.