User Input: Data is initially captured through user-facing web applications and forms.
Processor: Servers receive this request, process the underlying business logic, and format the raw data.
Storage: Finally, the database acts as the permanent, structured repository where data can be accessed, updated, or deleted
Press enter or click to view image in full size
The Foundational Data Pipeline
SQL vs. NoSQL Paradigms
Two Storage Categories: Data storage engines are divided by how they structure information: rigid tabular formats (SQL) and flexible document formats (NoSQL).
SQL Examples: Traditional databases like Oracle, MySQL, PostgreSQL, and MS SQL Server act like rigid filing cabinets.
NoSQL Examples: Modern alternatives like MongoDB, Redis, and Cassandra utilize dynamic, interconnected document folders.
Press enter or click to view image in full size
SQL vs. NoSQL Paradigms
The Constraints of Relational SQL
Rigid Schemas: Every record in a SQL table is forced to follow exact, predetermined structural rules without exception.
Tabular Structure: All information must fit strictly into predetermined columns and rows.
Complex Relations: Because tables are tightly linked, retrieving complex data requires writing long, intricate SQL queries
Press enter or click to view image in full size
The Constraints of Relational SQL
MongoDB’s Flexible Document Architecture
Open Collections: Instead of relying on rigid tables and rows, MongoDB groups independent documents into “collections”.
JSON-Style Formats: Data is natively stored in key-value pairs, which seamlessly mirrors the exact data structures used by modern web applications.
Non-Relational Speed: Because documents stand alone, there is no need for complex table joins; this makes fetching data incredibly fast and allows for very short queries
Press enter or click to view image in full size
MongoDB’s Flexible Document Architecture
Translating the Database Taxonomy
New Terminology: MongoDB shifts away from tabular naming conventions toward object-oriented terminology while keeping similar foundational concepts.
Table becomes a Collection.
Row becomes a Document.
Column becomes a Field.
Table Joins are replaced by Embedded Documents.
Press enter or click to view image in full size
Translating the Database Taxonomy
The Schema Flexibility Test
High Friction in SQL: Altering a SQL database requires modifying and rebuilding the entire column structure for all users, which creates risk and potential database downtime.
Zero Friction in MongoDB: Because MongoDB does not enforce a fixed schema, documents in the exact same collection can have entirely different fields. For instance, a “Married: Yes” field can be added instantly to a single user’s document without affecting any other records in the database, allowing applications to evolve continuously
Get Mohit Pratap Mehra’s stories in your inbox
Join Medium for free to get updates from this writer.
Press enter or click to view image in full size
The Schema Flexibility Test
The Underlying Engine (JSON vs. BSON)
JSON (JavaScript Object Notation): This is a text-based format primarily used for Web APIs and data transfer, powering 90% of web communication. However, it carries a slower processing overhead.
BSON (Binary JSON): MongoDB developers interact using familiar JSON syntax, but the system secretly compiles it into BSON. BSON is a highly compressed, lightweight, and ultra-fast binary format (0s and 1s) that serves as MongoDB’s hidden storage engine
BSON Expanding JSON’s Limitations
Standard JSON Limits: Standard JSON is restricted to basic types like Strings, Booleans, Arrays, Nulls, and Objects.
Extended Enterprise Capabilities: BSON introduces complex data types critical for enterprise-grade applications, drastically improving search capabilities and database precision. These extended types include 32-bit/64-bit Integers, Doubles, Regular Expressions, Timestamps, Dates, and Object IDs.
Press enter or click to view image in full size
BSON Expanding JSON’s Limitations
Infinite Scale Through Sharding
Horizontal Scaling: MongoDB avoids the need for expensive hardware upgrades by using a horizontal scaling technique called Sharding.
Automated Overflow: When a server reaches its maximum capacity, the database automatically distributes the massive datasets across parallel servers (moving data from Bucket A to Buckets B and C), allowing for effortless growth.
Press enter or click to view image in full size
Infinite Scale Through Sharding
Disaster Recovery and Replication
Built-in Bulletproofing: MongoDB ensures continuous operation through automated disaster recovery.
Replica Sets: The database continuously clones its data to secondary replica sets located in completely separate geographic locations. This guarantees zero downtime, even in the event that the primary server is completely destroyed.
Press enter or click to view image in full size
Disaster Recovery and Replication
Native Support for Modern Data Challenges
Geospatial Data: The database has native support for storing, indexing, and seamlessly querying location-based coordinates.
Big Data Analytics: MongoDB integrates effortlessly with complex data analytics tools and pipelines, providing deep enterprise intelligence.
Open Source Community: The architecture is free, community-driven, and supported by a massive global network of developers
Press enter or click to view image in full size
Native Support for Modern Data Challenges
Real-World Application (Stock Market Engine)
High-Frequency Analytics: Financial applications rely on MongoDB to handle millions of fluctuating shares per second.
The Power Trio: This real-time capability is powered by the combination of lightning-fast BSON writes, horizontal sharding to manage high traffic, and flexible schemas that can adapt to constantly varying financial assets
Press enter or click to view image in full size
Real-World Application (Stock Market Engine)
The Next Phase (Building a CRUD System)
Operational Commands: The final outlines the four dynamic operations required to master MongoDB data manipulation (CRUD):
Create: Inserting new document collections.
Read: Fetching and searching complex BSON data.
Update: Modifying schemas and data structures on the fly.
Delete: Safely removing irrelevant or outdated documents.
