docker-compose-spring-boot-postgres

A Java Spring and PostgreSQL project with CRUD API operations, containerized setup, OpenAPI-compliant documentation, pre-filled data, and Python Locust load tests for high concurrency simulation in Dockerized environments.

Target

1. 請實作一個 Java Spring Framework project 可對 PostgreSQL 做存取 POST /item GET /item?id=** Reference : https://github.com/bezkoder/docker-compose-spring-boot-postgres

2. https://swagger.io/specification/ 請參考 OpenAPI 3.0 的規範 實作輸出 API Specification https://springdoc.org/

3. 在 Database 內存放上萬筆資料

4. 使用 python locust 對 GET API 做壓力測試 模擬五百位使用者同時呼叫, 產出測試報告 Reference: https://locust.io/

Step 0. Init

Load parameters into shell environments

• source .env

Step 1. Start Demo App

• docker compose up -d

Make sure using docker ps command to check, app, postgres, python, 3 container is up

Demo App Home Page: http://localhost:8080

Step 2. Test POST/GET Method

• docker compose exec python-env python ./script/tests_post_get_method.py

Result

Step 3. Check Database Table Content

3.1

• docker compose exec postgresdb psql -U $POSTGRESDB_USER -d $POSTGRESDB_DATABASE -c "SELECT t.* FROM public.items t LIMIT 5;"

After step 2., it should have 5 data in the table.

Result

3.2

Truncate table

• docker compose exec postgresdb psql -U $POSTGRESDB_USER -d $POSTGRESDB_DATABASE -c "TRUNCATE TABLE items;"

Step 4. Check OpenAPI 3.0

swagger-ui

• URL: http://localhost:8080/swagger-ui/index.html

Result

api-docs

• URL: http://localhost:8080/v3/api-docs

Result

Step 4. Insert 10K Data Into Database

• Python Method with multi-processing
  • docker compose run python-env python ./script/insert_data_python.py

Result

• Or you can use helper script Import Table to save your time
  • ./script/import_table.sh

Step 5. Check Table Count

• docker compose exec postgresdb psql -U $POSTGRESDB_USER -d $POSTGRESDB_DATABASE -c "SELECT COUNT(*) FROM public.items;"

Result

Step 6. Locust Stress Test

For CLI

• docker compose exec python-env /bin/bash -c "locust -f ./script/locust_items_test.py --headless --users 500 --spawn-rate 100 -H http://app:8080 --run-time 10m"

For WebUI

• docker compose exec python-env /bin/bash -c "locust -f ./script/locust_items_test.py --users 500 --spawn-rate 100 -H http://app:8080 --run-time 10m"
• URL: http://localhost:8089

Locust Stress Test Conditions and Results

• Machine: 2022 MacBook Air M2
• CPU: 8-core
• RAM: 24 GB
• Operating System: macOS 13.4.1
• Test Conditions:
  • Users: 500 simulated users
  • Spawn Rate: 100 users/sec
• Results:
  • Requests Per Second (RPS): 164.8
• Note: Performance metrics are specific to the mentioned machine and setup, and may vary on different hardware or configurations.

500 User Statistics

500 User Chart

Other reports can be found in here: ./doc/locust_500_user_report/

Remove All

• docker compose down --rmi all

Helper Scripts

Some helper scripts to make life easier

• Load Database Table: ./script/import_table.sh
• Export Database Table: ./script/export_table.sh
• Simple Test POST Script: ./script/tests_post_method.sh
• Simple Test GET Script: ./script/tests_get_method.sh