Parasoft
Search
Parasoft Concerto

Load & Performance Testing

  • Validate functionality and performance under various loads
  • Simulate extreme loads and/or test from different locations
  • Expose concurrency issues in multithreaded applications
  • Define and enforce expected Quality of Service (QoS) metrics for SLAs
  • Test using existing Parasoft tests, JUnit tests, or lightweight socket-based components
Evaluation More Info eKit
Free Trial More Info Download eKit

For load and performance testing, Parasoft enables teams to reuse their existing test cases to verify application performance and functionality under heavy load.

Performance and Stress Testing

Existing functional tests are leveraged for load testing, removing key barriers to performance monitoring. The solution not only monitors the server's response rate with the specified number and mixture of simultaneous requests, but also verifies whether functionality problems occur under load.

Prebuilt scenarios can be used to verify robustness and scalability. You can easily customize these scenarios to use different test cases, load levels, load distributions, and so on. You can also distribute virtual users across remote server machines to simulate extreme loads and/or test from different locations.

Support is also provided for load testing non-Parasoft components such as JUnits or lightweight socket-based components. This provides teams an integrated solution for their various load testing needs.

Monitoring Tools

To help you collect network information and system performance data during load testing, Parasoft's solution provides built-in and extensible monitoring capabilities. For instance, supported monitors include perfmon, SNMP, rstat, WebSphere, WebLogic, JBoss, Tomcat, and remote monitors.

The solution's extensible and deployable monitoring framework allows you to extract just about any set of metrics from the system that is being subjected to load, graph and correlate those metrics visually to identify causes of performance issues, and apply expected quality of service policies on those metrics to get a high-level view of the system's compliance with performance and reliability requirements.

Concurrency Testing

Parasoft's load testing can also uncover concurrency issues in applications. For instance, if someone writes a JUnit test case for classes in the team's Java code, we can simulate threads accessing the code simultaneously to expose issues. This is increasingly important as today's applications are more and more multi-threaded and concurrent. In addition to applying load tests to uncover possible concurrency issues, Parasoft also provides advanced static analysis to uncover or prevent potential concurrency or deadlock issues.

Expected Quality of Service Management

Parasoft's solution defines and enforces expected Quality of Service (QoS) metrics that are important for setting and measuring SLAs (Service Level Agreements) during development and QA. For example, you can have the load test marked as a failure if the execution time is measured above 200 miliseconds. You can define success metrics before testing and set performance thresholds. Choose from a broad array of preconfigured metrics or define custom ones for your specific needs.

Continuous Load Testing

Parasoft solution's Load Test Continuum helps your incorporate performance testing throughout the SDLC and run performance tests continuously—so you can discover and resolve issues before they become a problem. It helps teams:

  • Eliminate SDLC uncertainty by performing load tests early and in a continuous automated fashion.
  • Track performance of services/applications throughout the SDLC, thus discovering and resolving issues before they become a problem.
  • Establish a QoS (Quality of Service) baseline for performance then detect deviations from that baseline as soon as they happen.

Service Virtualization for Performance Testing

Performance test engineers need to test against realistic transaction performance from dependent applications (mainframes, third-party services, databases, etc.). However, such dependent apps are often difficult or costly to access for performance testing. And when access is feasible, testers usually have a limited degree of control over how the dependent apps behave—and thus can test against only a limited range of behavior conditions.

With service virtualization, performance testers can

  • Test vs. realistic transaction performance.
  • Test constrained dependent resources without scheduling restrictions.
  • Test extensively—without access and transaction fees.
  • Test vs. a broad array of functional & performance conditions—with minimal setup.
  • Get the exact test environment they need, on demand.

To learn more about Parasoft's service virtualization and how it's applied in the context of load/performance testing, see our service virtualization page.

Copyright © 1996-2013 Parasoft   T: 888-305-0041   E: info@parasoft.com      [Site Map]     Privacy Policy