This repository contains scripts and data collected from running performance tests comparing an application written using monolithic architecture and microservices architecture.
Analysis (graphics and statistical tests) of the collected data is performed in a Collab Notebook.
Raw data is in the format of text files, collected from running WRK to load test the applications.
Data is processed using the results_parser.py script, which outputs data from WRK tests in JSON format.
The saved JSON files contain a list of JSON records, where each record represents the result of an individual WRK run.
Each JSON record contains the following fields:
avg: Average latency for the test
50%: 50th percentile latency (in ms)
75%: 75th percentile latency (in ms)
90%: 90th percentile latency (in ms)
99%: 99th percentile latency (in ms)
throughput: Requests/second for the test
Data is organized in two folders: monolith and microservices. Each folder contains data for the respective system.
The file names correspond to the test scenario executed for the system.
Test 1: Get pet owner details (1 instance of services)
Test 2: List Veterinarians (1 instance of services)
Test 3: Get pet owner details (3 instances of services)
Test 4: List Veterinarians (3 instances of services)