MongoDB Module 8: Security, Atlas Features, and Using MongoDB from Code

Step 1: Secure the deployment

Turn on authentication and create users with the least privileges they need, using built-in roles such as read and readWrite for a specific database. Encrypt connections with TLS, allow only known IP addresses or private networking, never expose the database port to the whole internet, and keep MongoDB updated. In Atlas, authentication, TLS, and encryption at rest are on by default, and you manage the IP access list and users.

use admin
db.createUser({
  user: "appUser",
  pwd: passwordPrompt(),
  roles: [ { role: "readWrite", db: "shop" } ]
})

Step 2: Protect connection strings and secrets

A connection string contains a password, so keep it out of source control. Read it from an environment variable or a secrets manager, use a dedicated user per application, rotate credentials when people leave, and never use the admin user in application code.

export MONGODB_URI="mongodb+srv://appUser:<password>@<cluster>.mongodb.net/shop"

Step 3: Connect from Python

PyMongo is the official Python driver. Create one client per application and reuse it, because clients manage a connection pool.

# pip install pymongo
import os
from pymongo import MongoClient

client = MongoClient(os.environ["MONGODB_URI"])
orders = client["shop"]["orders"]

orders.insert_one({"orderId": 100, "status": "pending", "total": 59.0})
for doc in orders.find({"status": "pending"}).sort("orderId", -1).limit(5):
    print(doc["orderId"], doc["total"])

pipeline = [{"$group": {"_id": "$status", "orders": {"$sum": 1}}}]
print(list(orders.aggregate(pipeline)))

Step 4: Connect from Node.js

The official Node.js driver follows the same ideas. Create the client once, connect, and close it when the program finishes.

// npm install mongodb
import { MongoClient } from "mongodb";

const client = new MongoClient(process.env.MONGODB_URI);
await client.connect();

const orders = client.db("shop").collection("orders");
await orders.insertOne({ orderId: 101, status: "pending", total: 12.5 });
const pending = await orders.find({ status: "pending" }).limit(5).toArray();
console.log(pending);

await client.close();

Step 5: Know Atlas Search, Vector Search, and backups

Atlas Search adds full-text search with relevance scoring, fuzzy matching, and autocomplete using search indexes. Atlas Vector Search stores embeddings and finds similar items, which supports semantic search and retrieval-augmented generation for AI applications. Atlas also provides continuous backups with point-in-time restore, monitoring, and alerts. Features and plan limits change, so check the current Atlas documentation and use a sandbox project while experimenting.

Step 6: Use MongoDB in data engineering

Common roles for MongoDB in a data platform: the operational store for an application, a source for pipelines that copy data into a data lake or warehouse with change streams or batch exports, and a serving layer for flexible, fast reads. Plan the flow from MongoDB to analytics early, because analytical queries over operational data can compete with application traffic, and flexible documents usually need flattening before they land in a warehouse.