I-Corps: Pain-free Database Administration via Workload Intelligence

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

Businesses are increasingly relying on database systems to support mission-critical applications. Consequently, running a database system has become a major operational cost for businesses as poor database performance can directly impact revenue and user experience. Database administrators (DBAs) are required to constantly monitor, diagnose, and rectify any performance decays. Unfortunately, the manual process of debugging and diagnosing database problems is extremely tedious and non-trivial. Rather than being caused by a single slow query, performance problems can be due to a large number of concurrent and competing queries creating compounded and hard-to-isolate effects. Sudden changes in request volume, query patterns, network traffic, or data distribution can cause previously abundant resources to become scarce, and the performance to plummet. This I-Corps team has developed a tool for assisting DBAs in quickly and reliably diagnosing performance problems in a database. By analyzing hundreds of statistics and configurations collected over the lifetime of the system, this tool's algorithm quickly identifies a small set of potential causes and presents them to the DBA. The goal of this project is to investigate the commercialization opportunities of this technology through customer discovery. Moreover, the team plans on locating and interacting with Beta users who are willing to use the proposed program and provide critical feedback on (i) what they like, (ii) what needs to be improved, (iii) what needs to be added, and (iv) what these customers are willing to pay.