Description

The task was to create DC Instance using the same instructions as the app performance toolkit: Postgres DB RDS (it was needed to check DB size), load project using project configurator, install google maps cloud and cloud core on the instance, create JMeter script to be executed with Taurus; run tests using Taurus to determine approximate maximum concurrency/throughput over 15 min test time frame (excluding ramp-up).

Goal

It was necessary to test the backend of the client's cloud addon, scanning for the workload. The client wanted to use the Atlassian framework provided for DC apps certification tests because the backend is hosted on Jira Service Desk DC. The key goal of the testing was to show end-users that the app can work under a heavy workload and is able to respond to many requests within a response time.

Solution

It was created a test-case list, conducted an investigation of a dynamic way to authorize JMeter tests and implement the first test; implemented java JWT lib for the JMeter (using the JwtHelper from the client's core). We used the lib in a JMeter test and implemented the first test and run it locally. Then we implemented tests using the test-case list, created DC instance equivalent to the prod, ran tests using the DC instance, different concurrency and tested duration parameters to find the edge.