IT ECOLOGY APPLICATION PERFORMANCE ASSURANCE (PRE-PRODUCTION) DISCIPLINES
Services offered before taking a new application or application version live include: Application Performance Testing, Application Profiling & Predictive Analysis and WAN Provisioning. These services ensure that applications have been streamlined to perform as best possibly with the least amount of costly resources whilst ensuring that the new application doesn’t negatively impact existing applications’ performance.
Application Performance Testing
Our skilled and highly-experienced consultants will plan and deliver a turn-key performance-testing project, from requirements gathering to scripting to execution and finally, full analysis of results including a detailed report highlighting the findings against SLA’s as well as exposing any areas where the system does not meet stated requirements with regards to Performance, Stability and/or Scalability.
Application Profiling and Predictive Analysis
Profiling allows you to get a view into an application’s network characteristics. It shows how a single user’s business transaction would perform across a given network and which component of the infrastructure is likely to become a bottleneck first. Predictive Analysis shows how transaction responsiveness is affected as certain network criteria (bandwidth, latency …) are adjusted and again shows which part of the application code is responsible for potential degradation. Since this form of testing does not rely on an expensive test environment and is relatively quick to perform, it should be done prior to performance testing.
Wan provisioning makes use of packet captures from the predictive analysis and models a network link which is traversed by multiple users transacting against an application with multiple business processes at specified frequencies. As various network criteria are changes in the model, the predicted response times for application steps are updated so that the best possible scenario can be found for planning into production.
Architecture verification includes verification of application and infrastructure architecture. The former verifies for example if a widget should be on the client or server to achieve optimal performance. The latter focusses on less obvious server infrastructure configurations and their performance impact. These may include resource allocations on JAVA virtual machines and how many JVMs are required to sustain a certain user count and so on. Other scenarios may include verifying synchronous vs. asynchronous database replication or checking data stability vs. performance aspirations during various failover conditions. Such verification is achieved with carefully designed performance testing test cases and is usually required when little information is available as to the best performing architecture design.