¹IBM T.J. Watson Research Center, 19 Skyline Drive, Hawthorne, NY10532, USA

Correspondence: Wim De Pauw,

Received 8 January 2009; Revised 21 February 2009; Accepted 24 February 2009.

Abstract

Stream processing is a new and important computing paradigm. Innovative streaming applications are being developed in areas ranging from scientific applications (for example, environment monitoring), to business intelligence (for example, fraud detection and trend analysis), to financial markets (for example, algorithmic trading systems). In this paper we describe Streamsight, a new visualization tool built to examine, monitor and help understand the dynamic behavior of streaming applications. Streamsight can handle the complex, distributed and large-scale nature of stream processing applications by using hierarchical graphs, multi-perspective visualizations, and de-cluttering strategies. To address the dynamic and adaptive nature of these applications, Streamsight also provides real-time visualization as well as the capability to record and replay. All these features are used for debugging, for performance optimization, and for management of resources, including capacity planning. More than 100 developers, both inside and outside IBM, have been using Streamsight to help design and implement large-scale stream processing applications.