Load Testing For Crypto Currency
**ABOUT THE CRYPTO CURRENCY PROJECT** Eclipton is a distinctive ecosystem utilizing Blockchain technology, facilitating social networking, freelancing, cryptocurrency trading, and blogging. A notable feature of this platform is its ability to function effectively.
**REQUIREMENTS**
- The client aimed to launch a new version for their 5,000 users.
- They chose to undergo Performance Testing to conduct stress tests with 5,000 concurrent users.
- The target benchmark for business transactions was set at three seconds during peak load.
**PROBLEMS AND CHALLENGES**
- During the script preparation phase, we encountered a significant challenge: the addition of AWS Cognito authentication servers prevented us from adding multiple users.
- For KYC verification, the system only accepts real Aadhaar or PAN card details, making it difficult to incorporate authentic Aadhaar card information for all 5,000 users.
- The blog page continued to load when we checked the UI after executing with 2,000 users attempting to create a free or paid blog.
- Tokens expired after two minutes.
- After entering the required information to create a gig and hire talent, the page remained unresponsive.
**OUR SOLUTIONS**
- Rather than registering 5,000 new users, we opted to reuse 4 users multiple times to navigate the AWS Cognito authentication issue.
- For KYC verification, we collaborated with the team to create 4 new users and approved KYC from the backend.
- We communicated the loading issue to the development team, and upon re-running the test with 2,000 and 5,000 users, we observed that it successfully loaded and displayed the number of blogs.
- We manually inserted a request to redirect to the appropriate page after creating a gig and hiring talent.
- The development team implemented necessary changes, allowing us to successfully execute the test a second time, demonstrating that the application can handle a load of 5,000 users.
**BENEFITS TO THE CLIENT**
- The application is now capable of supporting 5,000 users without errors like 500 while users engage in various scenarios.
- We achieved an average response time of under 3 seconds.