Performance Visualization of a Distributed System: A Case Study

Authors:
Hank Jakiela
Affiliations:
-
Venue:
Computer
Year:
1995

Citing 10
Cited 1

Parallel computer systems: performance instrumentation and visualization

Parallel computer systems: performance instrumentation and visualization
The visualization of parallel systems: an overview

Journal of Parallel and Distributed Computing - Special issue on tools and methods for visualization of parallel systems and computations
Recent developments and case studies in performance visualization using ParaGraph

Proceedings of the workshop on performance measurement and visualization on Performance measurement and visualization of parallel systems
Glyphmaker: Creating Customized Visualizations of Complex Data

Computer
Rdb (2nd ed.): a comprehensive guide

Rdb (2nd ed.): a comprehensive guide
Visualization and Debugging in a Heterogeneous Environment

Computer
Traceview: A Trace Visualization Tool

IEEE Software
Visualizing the Performance of Parallel Programs

IEEE Software
MulTVision: A Tool for Visualizing Parallel Program Executions

Proceedings of the US/Japan Workshop on Parallel Symbolic Computing: Languages, Systems, and Applications
Strata-various: multi-layer visualization of dynamics in software system behavior

VIS '94 Proceedings of the conference on Visualization '94

Algorithm Visualization For Distributed Environments

INFOVIS '98 Proceedings of the 1998 IEEE Symposium on Information Visualization

Quantified Score

Hi-index	4.10

Visualization

Abstract

Distributed systems are increasingly being used to handle large, business-critical on-line applications that demand guaranteed, quick end-user response times. However, tools for managing performance in distributed systems are not as mature as tools for single systems, and improved ways of handling system-wide performance problems are needed to reduce the cost and complexity of building and managing distributed systems. This case study describes the use of a simple visualization technique to help resolve difficult performance problems in a large distributed system. The use of application-level trace information provides a high-level overview of system behavior that complements the information provided by lower level and component-specific performance tools. Response time problems caused by lock contention, queueing delays, and soft faults can be easily identified with this technique.