[DB Seminar] Fall 2015: Zeyuan Shang

Finding real-world applications and workloads is the bane of every database researcher. To overcome this problem, we present the Carnegie Mellon Database Application Catalog (DBAC). The DBAC finds database applications from on-line source code repositories (e.g., GitHub) and then automatically installs them in a virtual machine sandbox. It then submits requests to the application that cause it to interact with a database. The DBAC monitors all database queries the application invokes and stores them in a central repository for analysis.

In this talk, we present the design and implementation of the DBAC system. The current catalog contains over 3000 applications. We discuss how the system is able to automatically deploy and execute these applications. We also present some initial analysis of the characteristics and complexity of these workloads. We conclude with our plan for making these applications available to other researchers for benchmarking and expand the breadth of the catalog.

Bio:
Zeyuan Shang is a visiting researcher with the Carnegie Mellon Database Group. He is completed his undergraduate degree in Computer Science at Tsinghua University.