Load Test of Banking Domain Solution


A North American company engaged Newgen to perform the Load Test of their product. The product was a solution for managing the transactions done on ATMs like Cheque deposit, Cash Deposit etc. The Client was looking to maintain QoS (Quality of Service) for their release of entire product line requiring an end to end automated solution for load testing of the product and monitor major Key Performance Indicators like CPU, Memory, Disk I/O Usage etc. Newgen team followed Agile methodology to keep up with the deployment targets.


The tool needs to be generic and configurable. The key challenges were:

  • Simulating 10,000 concurrent ATMs and dividing the ATMs according to the defined transaction load.
  • Designing framework in such a manner through which test engineer has the complete control on number of transactions and type of transaction to be done in a specified duration.
  • Capturing major KPI’s like CPU, Memory, Disk I/O, Network I/O Usage.
  • Simulating the identified test scenario of all the independent modules of an application in a single script.
  • Capturing Logs from all the Servers like Database Server, IIs Server automatically after each test run and storing along with the final Consolidated Report
Case Study Hurdles


In Initial stage Newgen studied the requirements and then testing objectives were defined in agreement with client. To keep software testing cost low open source tools were evaluated and JMeter was selected as best Fit. For Capturing the required KPI’s a JMeter Plugin PerfMon was used. Test environment required for ATM load was validated and setup at client base.

Test scenarios were identified and scripts were designed through JMeter tool. In parallel a utility was developed using c# to completely automate the test automation process steps from test execution activity till Report generation.

The report not only covered key Knowledge Performance Indicators but also included several additional KPI’s like transaction Per second, Hits per second graphs that helped locating the bottlenecks in the application which proved to be beyond the expectation of the customer.


  • Generating centralized final test report helped in evaluating the performance and finding bottlenecks in the application.
  • Automating the entire test helped in frequent verification of application confirm the adherence with specified benchmarks.
  • Usage of open source tool for load test provided huge savings in terms of license cost.
  • These tests significantly reduced risk of system failure and poor performance.


Tools and Technologies

JMeter Tool, JMeter Plugin PerfMon, C#, Javascript