Contact us
Performance Testing for a bidding App

Overview

Weblanza made a web-based bidding app for the government of Qatar so that they could run public auctions for cars that had been taken by police. People who used the platform could bid on items in live web auctions in real time.

However, the system had trouble handling a lot of people during auction events with a lot of bids. The client wanted the platform to be able to handle 5,000 bidders at the same time and 9,500 deals per hour without slowing down. Weblanza worked with Testvox to do structured speed testing and optimisation on the app to make sure it could meet these needs.

Challenges

Before working with Testvox, Weblanza had a lot of major performance issues, such as:

  • A lot of extra traffic During sale days, thousands of users logged in at the same time, which caused traffic to suddenly spike. Things moved very slowly on the system, and some users were not able to make bids.
  • Failures of Applications Under Load The goal was for 5,000 users to be able to use the app at the same time, but it started to fail when real-time users went over 500. It would become unsteady after 5–10 minutes of heavy use.
  • Unstable servers and high costs The cloud-based computer wasn't set up to handle a lot of traffic. This caused operational costs to go up and made things less stable when the number of users grew.

Our Solution

Testvox made a structured plan for speed testing to find problems and give the client the confidence to grow.

  1. A method for testing with gradual loads

    2. We started testing with 500 users and slowly added more until we had 1,000 users and then 2,000 people. When errors showed up when the server had a lot of traffic, we closely watched the server metrics to find the reasons. The production team was told about the findings so that they could be fixed right away.

  2. Testing Again After Optimisation

    3. After Weblanza made the changes that were suggested, we ran the tests again, this time with up to 5,000 people at the same time. Carefully keeping an eye on the server's performance, more tweaks were made until steadiness was reached with acceptable response times.

  3. Tough analysis and monitoring

    4. We tested for load, stress, and endurance with JMeter. A lot of attention was paid to how much CPU, memory, and network speed were used. The tests kept going until the CPU usage levelled off below reasonable levels and performance as a whole got a lot better.

  4. Testing in a production environment

    5. The app was already live, so testing had to be done in the production setting, where real users wouldn't be affected. To deal with this, testing was done when no one was working, so it wouldn't affect people who were bidding at the time.

Result

Work that stays the same at 5,000 users

Now, the app can handle 5,000 real-time bidders with stable server behaviour and reaction times.

9,500 transactions per hour were completed.

The system can handle the necessary 9,500 transactions per hour thanks to workload modelling and scripting that works better.

Delivered in two weeks

Testvox finished coding, execution, analysis, and problem resolution in just two weeks, even though they had to meet a tight deadline.

Final Thoughts

Testvox helped Weblanza turn a system that wasn't working well into a stable, high-performing bidding platform by using organised load testing and careful server analysis. The app now works perfectly during auctions with a lot of bids, so thousands of users can join without any problems. Today, the platform is being used successfully for online car auctions, meeting standards for both performance and dependability.

Related Resources