WiSeDB: Cloud Workload Management via Machine Learning

Ryan Marcus
Brandeis University
Sofiya Semenova
Brandeis University
Olga Papaemmanouil
Brandeis University


Data management applications deployed on IaaS cloud environments must simultaneously strive to minimize cost and provide good performance. Balancing these two goals requires complex decision-making across a number of axes: resource provisioning, query placement, and query scheduling. While previous works have addressed each axis in isolation for specific types of performance goals, we present WiSeDB, a cloud workload management advisor service that uses machine learning techniques to address all of the problem for customizable performance goals.

Supervised learning approach for batch workloads
Reinforcement learning approach for online workloads