HP Service Virtualization: Answering the Challenge of Performance Testing

Bad performance test designs can generate load-on components that are out of scope, resulting in very inefficient performance testing practices.  This means that a performance tester will spend time scripting or focusing on performance issues with a component that is completely out-of-scope or unchangeable.  Additionally, performance testers face challenges that pose a serious threat to the success of application development efforts.  If there is no plan for overcoming these issues, the result will be project delays, budget overruns, or even total system failure in production. 

 

HP Service Virtualization addresses these challenges and allows us to understand the components that shouldn’t be distractions.  This way, users can focus on performance characters of the code and components that are most important to the project.   

 

man at computer.png7 benefits of Service Virtualization for performance testing:

  • Avoid accidental overload.  If you are unaware of back-end system calls during a load test, it’s very easy to unintentionally send high-throughput traffic (similar to a denial-of-service attack) to an external system to which you have no permission to be testing.  Virtualizing third party or external service calls is a safer solution and will help keep partners safe, too.
  • Begin performance testing earlier in the lifecycle.  Contemporary agile methodologies require performance testing to be the first components built during the first sprint of development, before any other components have been completed.  HP Service Virtualization is a critical enabler to allow users to get started with performance testing from the first sprint.
  • Prevent late-night performance testing schedules.  Performance testing environments are often only available for load testing in off-peak hours, when no other critical business functions or transactions will be disrupted.  Virtualizing those database or mainframe sources enables users to run tests during normal business hours, avoiding the additional overtime or after-hours costs. This also creates schedule alignment with development.
  • Overcome budgetary limitations for lab resources.  When performance lab budget is tight, there are always some secondary components for the system that are nice to have, but not absolutely required; by virtualizing those components of the systems you can eliminate costs. Performance test accuracy will be preserved for the main components of the application system that are proprietary, self-built, customized and most important.
  • Avoid performance test work stoppage.  At times during a performance test, users may hit a minor defect with a sub-component of the system, which can hinder testing the performance of the full end-to-end transaction.  By quickly virtualizing erroneous components or services, performance testing can continue as if the defect or bottleneck was already resolved.  Project timelines and release schedules can often be met, with a critical patch or fix to be released shortly after the go-live date.
  • Improve performance testing analysis and prediction.  It is common to find performance tests that include very unpredictable external service response times, thus preventing any accurate extrapolation of test results.  By virtualizing those services consistently and with repeatable, predictable service emulation, you can repeat performance tests with exact conditions and greatly enhance the accuracy of your extrapolations and predictions of performance.
  • Reduce performance test complexity.  Performance test design for modern systems can become so complex with so many interdependent components that you might not know where to even start your performance test.  By virtualizing the majority of the sub-components, you can begin your testing with just a small subset of services or components and gradually increase the scope of your testing with each sprint.

service virtualization.png

 

Service virtualization allows users to virtualize components that shouldn’t be distractions and focus on the key performance characteristics of the code and components that are critical to the project.  Advancements in virtualization solutions can make it much easier to emulate and simulate portions of the application that have been unavailable during performance testing.  Consequently, the accuracy and integrity of performance measurements taken from highly virtualized applications are greatly improved. They are much more valuable to decision making earlier in the application lifecycle.

 

As modern performance practices face new challenges, the need to virtualize application dependencies is critical to success. There is real pressure to start performance testing earlier, to pair performance testers with developers, complete performance testing on time before release, and to improve the accuracy of results.  HP Service Virtualization is the right solution to respond to those pressures and enable the performance team to modernize, improve, and scale.  Integrated with HP Load Runner and Performance Center testing tools, it is a seamless solution for performance engineering and validation.  HP Service Virtualization directly impacts operational costs for performance testing and development, reduces pre-release project budgets/costs for performance efforts, and helps to ensure project timeliness.

 

Additionally, Service Virtualization integrates with HP Application Lifecycle Management (ALM) and HP Quality Center (QC) as a shared repository, simplifying execution of application lifecycle activities.  These activities leverage virtual services including, developer tasks, lab management, performance and functional test execution, and defect management.  Support for end-to-end testing of virtual services is enhanced with integration to HP testing solutions including: HP Unified Functional Testing (UFT), HP LoadRunner (LR), and HP Performance Center.  These platforms represent a virtualization of a wide breadth of service transports and protocols, from Web and Service Oriented Architecture (SOA) services to proprietary middleware technologies, mainframe technologies, and ERP, and B2B protocols.  Additionally, they support modeling data, performance, and network behavior including complex network models integrated from HP Shunra Network Virtualization data manipulation in virtual services, such as masking.

Have some great insight on Service Virtualization? Leave us a comment below and share your ideas! Or, join the conversation at @HPSoftwareALM.  Be sure to check us out at: HP Application Lifecycle Management and HP Service Virtualization: http://www.hp.com/go/sv.

 

Have some great insight on Service Virtualization? Leave us a comment below and share your ideas! Or, join the conversation at @HPSoftwareALM.  Be sure to check us out at: HP Application Lifecycle Management.

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the Community Guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
About the Author
Ferhan Kilical has over 20 years of experience in Information Technology with expertise in application development and testing as a Technica...


Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation