BENCHMARKING PARALLEL REQUEST PROCESSING SPEEDS ON AWS CLOUD SERVICES

O. Sihunov, L. Demkiv

Abstract


The article presents the results of testing the web application for conducting sports competitions in cloud services. Testing on different types of servers of the same price category under scenarios closest to real ones, in particular simultaneous data acquisition by many users, was carried out. The web application is written using the Spring Java framework, the Maven dependency builder, and the IntelliJ IDEA development environment. The project uses Scala and Gatling for testing. The program consists of two independent microservices, one of which is responsible for user registration and management, and the other for real-time data transmission. An important factor in testing is the choice of testing strategy.

 A scenario of a stable number of users over a certain period of time was chosen for testing the web application. Such a strategy is implemented by rampUsers(N) during M, which allows during a certain time M to call a total of N users. Performance tests with the number of users ranging from 100 to 600 users per second in steps of 100 users and performance tests of 500 users per second for an hour were performed for different types of instances. The parameters are defined: the stability of the application on a certain type of server, the speed of processing the request by the server for different numbers of users, and errors that occur during the operation of the application.

The most optimal type of servers for this application is determined, which gives the maximum ratio of performance and stability to the price for this type of tasks. Instances of the t3.large and r4.large types are best suited for the implementation of this project. Among the advantages of r4.large is greater resilience to unexpected influxes of users thanks to faster memory and network.  However, t3.large is more balanced and stable, and the key advantage of t3.large is its significantly lower hourly price than r4.large.

Keywords: cloud services, instances, AWS EC2, Gatling, types of loads, test scenarios.


References


  1. Gao J. Cloud Testing-Issue, Challenges, Needs and Practice / J. Gao, X. Bai, W. Tsai // Software Engineering: An International Journal (SEIJ). -Vol. 1, No. 1. -P. 9-23.
  2. Lnenicka M. Classification and Evoluation of Cloud-Base Testing Tools: The Case Study of Web Application Security Testing/ M. Lnenicka, J. Capek// Acta Informatica Pragensia, 2018, 7(1), 40–57. DOI: 10.18267/j.aip.113
  3. Liu W. Performance Test and Improvement of Computer Network Virtualization Software in Cloud Computing Environment//Volume 2022 | Article ID 6965880 https://doi.org/10.1155/2022/6965880
  4. Li H. Research on Cloud Performance Testing Model/ Li H., Li X., Wang H., Zhand J. //Conference: IEEE 19th International Symposium on High Assurance Systems Engineering, 2019. DOI:10.1109/HASE.2019.00035




DOI: http://dx.doi.org/10.30970/eli.20.4

Refbacks

  • There are currently no refbacks.