[TIL] Converting MySQL -> MongoDB (NoSQL)

📌 Issues encountered

Issue #1

When querying the entire chat history from MySQL, the query took too long. When testing with Postman, it took around 2.xx seconds.

Issue #2

Adding MongoDB connection.

📌 What I tried

Issue #1

• There were two tables to join with the Chats table in MySQL, and the part where the entire chat history was queried required a lot of query executions and processing to manipulate the returned objects in the service layer.

• Therefore, I decided to change the Chats table in MySQL to MongoDB, and created two schemas: Chats schema and Messages schema.

• Chat-related data is stored in MongoDB, while the rest is stored in MySQL, so I added MySQL data to the Chats schema to minimize the parts that fetch data from MySQL while performing CRUD operations on Chat.

• The Chats schema includes fields such as title, is_sold, product_id, members, members_nickname, and address.

• After actually changing the Chat part to MongoDB, the query time decreased from 2.xx seconds to around 600 ms.

Issue #2

1. In utils/mongodb.js, I defined the part that connects to MongoDB as follows:

const mongoose = require("mongoose");
const { mongodb } = require("../config/config.js");

const db_url = `mongodb+srv://${mongodb.id}:${mongodb.password}@${mongodb.cluster}/${mongodb.db}?retryWrites=true&w=majority`;

const connectToMongoDB = async (app) => {
  mongoose
    .connect(db_url)
    .then(() => {
      console.log("MongoDB connected")
    })
    .catch((err) => {
      throw err;
    });
};

module.exports = connectToMongoDB;

1. In app.js, I imported MongoDB as follows:

const connectToMongoDB = require("./utils/mongodb");
connectToMongoDB(app);

1. In the schema folder, I defined the Chats and Messages schemas and imported them in the respective repositories where these schemas are used.