Date of Award
6-2025
Document Type
Thesis
Publisher
Santa Clara : Santa Clara University, 2025
Degree Name
Master of Science (MS)
Department
Computer Science and Engineering
First Advisor
Shiva Jahangiri
Abstract
Modern database management systems must balance high throughput with low-latency query execution across diverse and dynamic workloads, where queries vary significantly in complexity, resource demands, and latency sensitivity. This challenge is further compounded in multi-tenant environments, where the system must concurrently serve a wide range of query types while adapting to workload fluctuations and ensuring stable performance. Achieving this balance requires intelligent scheduling mechanisms that can effectively classify, prioritize, and allocate resources to queries in a manner that aligns with both user expectations and system-level performance goals.
This thesis surveys both academic and industrial advancements in workload and resource-aware query scheduling techniques designed to meet varying performance requirements. Drawing inspiration from heuristic scheduling strategies and recent developments in industrial systems, the thesis presents the design and implementation of a dynamic workload management module for AsterixDB. The proposed scheduler accommodates both user- and system-level requirements by enabling semantic query classification, user-defined prioritization, and fairness-aware scheduling policies tailored to different workload scenarios. To evaluate its effectiveness, we simulate three representative workload patterns under high query concurrency and conduct a comprehensive experimental analysis. The results demonstrate the importance of fine-grained query characterization, memory-aware scheduling, and feedback-driven control in supporting responsive and multi-tenant workload management in highly concurrent database systems.
Recommended Citation
Shi, Hongyu, "Dynamic Workload Management for Highly Concurrent Database Management System" (2025). Computer Science and Engineering Master's Theses. 49.
https://scholarcommons.scu.edu/cseng_mstr/49
