🍃 MongoDB Magic — From Zero to Rock-Solid in Minutes! 🚀

🚀 From Chaos to Order: How Mongock & Friends Keep Your DB in Shape Automatically!

☁️ The “Cloudyrock” Mystery

You might have seen this in a Spring Boot project:

@ChangeLog

public class DatabaseChangelog {

  

    @ChangeSet(order = "001", id = "createIndex", author = "anand")

    public void createIndex(MongoTemplate mongoTemplate) {

        mongoTemplate.indexOps("users")

                     .ensureIndex(new Index().on("email", Sort.Direction.ASC).unique());

    }

}

And wondered:

💭 "Wait… where’s my CREATE INDEX script? Who’s applying this? And why is it happening without me touching the DB?"

That’s Mongock in action — made by the Cloudyrock team.
It’s basically the database time machine for Java projects.
No flux capacitors, just clever scanning, execution, and bookkeeping.

🔍 How It Works (From Start to End)

1️⃣ Scan Your Change Logs

You specify in application.properties:

mongock.change-logs-scan-package=com.javabeanbag.changelog

Mongock starts with Spring Boot and scans this package for classes annotated with @ChangeLog.

2️⃣ Run Only What’s New

Inside each @ChangeLog, you define @ChangeSet methods.

• id → Unique name for the change
• order → Execution order
• author → Who wrote it

Mongock stores the executed changes in a special collection/table (e.g., mongockChangeLog in MongoDB).

Before running a change, it checks:
“Is this id already in my history table?
If yes → skip
If no → run and record it.”

This prevents duplicate execution and “index already exists” errors.

3️⃣ Apply Changes

• For MongoDB → Uses MongoTemplate or MongoDatabase to run commands
• For MySQL/PostgreSQL (JDBC mode) → Runs SQL via JDBC

✅ Yes — Mongock can work with MySQL, but for SQL databases, Liquibase/Flyway are more common.

📌 Example — Adding an Index in MongoDB

@ChangeLog(order = "001")

public class UserIndexes {

    @ChangeSet(order = "001", id = "createEmailIndex", author = "anand")

    public void createEmailIndex(MongoTemplate mongoTemplate) {

        mongoTemplate.indexOps("users")

                     .ensureIndex(new Index().on("email", Sort.Direction.ASC).unique());

    }

}

📍 What Happens in the DB

After the first startup, Mongock creates a special collection in MongoDB:

mongockChangeLog

{

  "_id": ObjectId("66bb98f4134a6e24f0e91234"),

  "changeId": "createEmailIndex",

  "author": "anand",

  "timestamp": ISODate("2025-08-13T10:20:00Z"),

  "state": "EXECUTED",

  "executionMillis": 120,

  "executionHostname": "dev-machine",

  "executionId": "f27b81f2-2341-4d8f-8d4b-23bb53f1c9ae"

}

📌 Next Startup: Mongock checks changeId → “Already exists” → Skips execution.

💡 Example — Seeding Initial Data in MySQL

@ChangeLog(order = "002")

public class InitialData {

    @ChangeSet(order = "001", id = "insertAdmin", author = "anand")

    public void insertAdmin(Connection connection) throws SQLException {

        try (PreparedStatement stmt = connection.prepareStatement(

             "INSERT INTO users (id, username, role) VALUES (1, 'admin', 'ADMIN')")) {

            stmt.executeUpdate();

        }

    }

}

⚔️ Mongock vs Liquibase vs Flyway — Quick Battle Card

Feature	Mongock	Liquibase	Flyway
MongoDB Support	✅ Native	❌	❌
MySQL/Postgres	✅ (JDBC)	✅	✅
Java Code Changes	✅	❌ (SQL/XML)	❌ (SQL only)
Microservices Ready	✅	✅	✅
Best For	Mixed NoSQL+SQL	Complex DB	Simple SQL

😂 Life Without Mongock

• 🧠 Remembering which DB got which index
• 🖊 Writing “DB update” docs nobody reads
• 🏃‍♂️ Running manual scripts at midnight before deployment
• 🐛 Accidentally applying the same change twice

☕ Life With Mongock: Drink coffee → Deploy → Mongock: “Relax, I got this.”

🔄 Wrapping Up ☕

When working in microservices, keeping DB schemas in sync is like keeping cats in a bathtub — chaos 🐱💦.

Mongock (or Liquibase/Flyway) brings order:

• Tracks applied changes
• Avoids duplicates
• Works across environments without manual steps

💡 Rule:
MongoDB → Go Mongock
Pure SQL → Use Liquibase or Flyway