Contact us
Ensuring Scalable API Performance for a Leading Saudi Fintech Platform

Overview

People and companies can handle their money online more easily with OneAccount.sa, a fintech platform based in Saudi Arabia. That way, you can safely do business online, keep an eye on your funds, and get real-time information. This makes money more open and builds trust with users. Before putting the platform into use, the customer had to make sure that its basic APIs could handle real-world traffic loads and keep performance stable.

Problems

  • Checking how well an API works when it's busy It was important for the client that their APIs worked in the real world. As much as possible, reaction times and stability had to be the same at all the key places where sensitive financial transactions took place.
  • Making the most of what you have During times of high demand, it was very important to keep computer resources like CPU, RAM, and disc I/O at their best levels. This would keep things from getting worse and make sure that people had a good time when it was busy.
  • Locating and getting rid of snags During the first tests, issues such as restricted access and dependencies began to show up. The customer had to find these slowdowns right away and fix them in the server so that the platform could be used in real life.

Our Solution

Testvox made sure the platform was ready for production by setting it up for production, writing detailed test scripts, and running increasing load tests.

  1. Making the place ready

    2. We set up a different Google Cloud environment that was a copy of the client's production environment to make sure the tests were as real as possible. In order to get accurate information about performance, the logs and tracking tools were set up.

  2. Making tests and writing scripts

    3. To test traffic on the platform's most important routes, we used Apache JMeter to create scripts that worked like APIs. These lines were made to be used over and over, and they talked about all the important things.

  3. Making Load Models

    4. There were tests done at 30%, 50%, and 100% of the expected traffic to see how well the platform could handle more people. Each time through the tests, we looked at the data and made any changes that were needed.

  4. Observing and Writing Down

    5. Watching the CPU, memory, and disc I/O in real time showed that they were never used more than 75% of the time. This showed that the system could handle stress.

  5. Talks and suggestions

    6. We gave detailed data on response times, throughput, and success rates that showed all APIs met performance SLAs and could handle the expected number of transactions when they were going full speed.

Result

How Well It Works When Fully Loaded

The system was able to handle all the target activities even when it was nearly full, and response times stayed well within the SLA (<1 second).

High Performance Under Peak Traffic Loads

The platform was able to handle a lot of traffic during busy times because CPU, memory, and disc I/O use stayed well below 75%.

Easy to carry out

The platform was ready for production right away after the last tests showed that there were no major issues.

The power to grow and be ready for production

There are no more problems with the program, and it is now ready to be used by everyone.

Final Thoughts

Testvox did full API speed testing on OneAccount.sa to make sure it was ready to go. We helped them make sure that their platform could handle more people, use its resources more efficiently, and get rid of any problems that were stopping it from working. This gave them the confidence to launch the app and make sure that users had a smooth, safe experience. It is now possible for the platform to handle real-world traffic. This means that it will work well for everyone in Saudi Arabia.

Related Resources