Microservice Bad Smells Catalog

This repository contains the deployable implementation and seed database for an online Microservice Bad Smells Catalog. The catalog is associated with the paper:

Yongchao Xing, Zhiying Tu, Dianhui Chu, Weipan Yang, and Yiming Lv. Catalog, Impact, and Evaluation of Microservice Bad Smells: A Systematic Literature Review. Software: Practice and Experience, 56(4):372–423, 2026. DOI: 10.1002/spe.70053.

The system provides a web-based knowledge base for browsing, searching, and sharing microservice bad smell knowledge. Each smell entry records its aliases, description, detection method, consequences, causes, examples, classification, evidence score, literature sources, and related smells.

Features

• Browse the complete microservice bad smell catalog.
• Filter smells by high-level categories, including internal design, communication and interaction, structure and infrastructure, decomposition, security, lifecycle management, and team/technology.
• Search smells by name in the web interface.
• View detailed smell knowledge, including description, detection, consequence, cause, example, source references, and related smells.
• Copy smell information from the interface for reuse in research, documentation, or tool integration.
• Browse smell type information in table and tree-map views.
• Use the backend REST APIs to retrieve smell entries, smell types, and smell relations.

Repository Structure

.
├── catalog/
│   ├── dist/                         # Built frontend assets
│   │   ├── index.html
│   │   ├── css/
│   │   └── js/
│   ├── msbs-backend-8088.jar          # Spring Boot backend executable JAR
│   └── nohup.out                      # Example runtime log; can be ignored
├── sql/
│   ├── msbs_en.sql                    # Main microservice bad smell catalog database
│   ├── ts_auth_mysql.sql              # Auxiliary/example SQL files
│   ├── ts_config_mysql.sql
│   ├── ts_consign_mysql.sql
│   ├── ts_consign_price_mysql.sql
│   ├── ts_contacts_mysql.sql
│   ├── ts_food_mysql.sql
│   ├── ts_station_mysql.sql
│   ├── ts_train_mysql.sql
│   ├── ts_travel_mysql.sql
│   └── ts_user_mysql.sql
└── README.md

Note: If the archive contains a nested catalog/catalog/ directory, it is a duplicated deployment copy. The root-level catalog/dist/, catalog/msbs-backend-8088.jar, and sql/msbs_en.sql are sufficient for a normal deployment.

Technology Stack

Frontend

• Vue 3
• Vue Router
• Element Plus
• AntV G6
• umi-request
• citation-js

The provided frontend is a built deployment artifact under catalog/dist/.

Backend

• Java 8
• Spring Boot 2.7.14
• MyBatis Plus 3.5.2
• MySQL Connector/J
• Knife4j / Swagger support

The backend is provided as a runnable Spring Boot JAR: catalog/msbs-backend-8088.jar.

Database

• Recommended: MySQL 8.0+
• Main database script: sql/msbs_en.sql

The main catalog database contains:

• antipattern_info: smell entries
• antipattern_type: smell category/type hierarchy
• antipattern_relation: relations among smells
• user: basic user table used by the backend

The current seed data contains 69 smell entries, 23 type records, and 12 smell-relation records.

Quick Start

1. Prepare MySQL

Create a database:

CREATE DATABASE msbs_en DEFAULT CHARACTER SET utf8mb4 COLLATE utf8mb4_0900_ai_ci;

Import the main catalog data:

mysql -u root -p msbs_en < sql/msbs_en.sql

If you use MySQL 5.7 or another database version that does not support utf8mb4_0900_ai_ci, replace it with utf8mb4_general_ci or another compatible collation before importing the SQL file.

2. Start the Backend

Run the backend JAR and override the database configuration with your local database information:

java -jar catalog/msbs-backend-8088.jar \
  --server.port=8088 \
  --spring.datasource.url="jdbc:mysql://127.0.0.1:3306/msbs_en?useUnicode=true&characterEncoding=utf8&serverTimezone=Asia/Shanghai" \
  --spring.datasource.username="YOUR_USERNAME" \
  --spring.datasource.password="YOUR_PASSWORD"

After startup, the backend listens on:

http://127.0.0.1:8088

3. Start the Frontend

Serve the built frontend files:

cd catalog/dist
python3 -m http.server 8000

Then open:

http://127.0.0.1:8000

4. Configure Backend Address for the Frontend

The built frontend may contain a hard-coded backend address. If the frontend cannot access the backend, replace the backend URL in the built JavaScript files:

grep -R "47.92.224.21:8088" -n catalog/dist/js

For local deployment on Linux, you can replace it with:

sed -i 's#http://47.92.224.21:8088#http://127.0.0.1:8088#g' catalog/dist/js/*.js

On macOS, use:

sed -i '' 's#http://47.92.224.21:8088#http://127.0.0.1:8088#g' catalog/dist/js/*.js

For production deployment, replace it with the public backend API address or rebuild the frontend from source using the desired API base URL.

Main API Endpoints

The frontend mainly uses the following backend APIs.

List all smell entries

GET /antipatternInfo/list

Example:

curl "http://127.0.0.1:8088/antipatternInfo/list"

List smells by category

GET /antipatternInfo/list/getByCategoryName?catagoryName={categoryName}

Example:

curl "http://127.0.0.1:8088/antipatternInfo/list/getByCategoryName?catagoryName=Internal%20Design"

The parameter name is intentionally written as catagoryName because this is the name used by the current frontend and backend implementation.

List related smells

GET /antipatternRelation/list/getRelatedAntipatterns?name={smellName}

Example:

curl "http://127.0.0.1:8088/antipatternRelation/list/getRelatedAntipatterns?name=Bloated%20Service"

List smell types

GET /antipatternType/list

Example:

curl "http://127.0.0.1:8088/antipatternType/list"

Data Model

antipattern_info

Stores the main smell knowledge entries.

Field	Description
id	Primary key
name	Smell name
aliases	Alternative names
description	Smell description
detection	Detection method or detection clue
consequences	Potential quality impacts
cause	Possible causes
example	Example scenario
sources	Literature sources, stored as JSON text
categoryName	High-level category
typeName	Fine-grained type
evidence	Evidence or impact score
isDelete	Logical deletion flag

antipattern_type

Stores category/type information.

Field	Description
id	Primary key
name	Type name
description	Type description
parentName	Parent type name; NULL for top-level categories
orderIndex	Display order
isDelete	Logical deletion flag

antipattern_relation

Stores relations between smells.

Field	Description
id	Primary key
AntiPatternName	Source smell name
relatedAntiPatternName	Related smell name
relation	Relation type, such as precedes or relates
isDelete	Logical deletion flag

Deployment Notes

Backend configuration

The backend JAR contains default database settings. For security and reproducibility, do not rely on embedded production credentials. Always override database configuration through command-line parameters, environment variables, or an external Spring Boot configuration file.

Example using an external configuration file:

java -jar catalog/msbs-backend-8088.jar --spring.config.location=file:./application-prod.yml

Example application-prod.yml:

server:
  port: 8088

spring:
  datasource:
    driver-class-name: com.mysql.cj.jdbc.Driver
    url: jdbc:mysql://127.0.0.1:3306/msbs_en?useUnicode=true&characterEncoding=utf8&serverTimezone=Asia/Shanghai
    username: YOUR_USERNAME
    password: YOUR_PASSWORD
  mvc:
    pathmatch:
      matching-strategy: ANT_PATH_MATCHER

mybatis-plus:
  configuration:
    map-underscore-to-camel-case: false
  global-config:
    db-config:
      logic-delete-field: isDelete
      logic-delete-value: 1
      logic-not-delete-value: 0

Frontend deployment with Nginx

Example Nginx configuration:

server {
    listen 80;
    server_name your-domain.example;

    root /path/to/catalog/dist;
    index index.html;

    location / {
        try_files $uri $uri/ /index.html;
    }
}

If the backend and frontend are deployed on different domains or ports, make sure CORS is enabled on the backend and the frontend API base URL points to the correct backend address.

Troubleshooting

Frontend shows ERR_FAILED, 502, or cannot load smell data

Check whether the frontend is still pointing to an old backend address:

grep -R "http://" -n catalog/dist/js | head

Then replace the backend URL with your deployed backend address.

Backend cannot connect to MySQL

Check:

• Whether MySQL is running.
• Whether the database msbs_en exists.
• Whether sql/msbs_en.sql has been imported.
• Whether the username and password are correct.
• Whether the JDBC URL uses the correct host, port, database name, and timezone.

SQL import fails because of collation

If your MySQL version does not support utf8mb4_0900_ai_ci, replace it:

sed -i 's/utf8mb4_0900_ai_ci/utf8mb4_general_ci/g' sql/msbs_en.sql

Then import the SQL file again.

Citation

If you use this knowledge base or system in academic work, please cite:

@article{xing2026catalog,
  title   = {Catalog, Impact, and Evaluation of Microservice Bad Smells: A Systematic Literature Review},
  author  = {Xing, Yongchao and Tu, Zhiying and Chu, Dianhui and Yang, Weipan and Lv, Yiming},
  journal = {Software: Practice and Experience},
  volume  = {56},
  number  = {4},
  pages   = {372--423},
  year    = {2026},
  doi     = {10.1002/spe.70053}
}