MongoDB

Before learning queries you need an environment — locally or in the cloud via Atlas.

Create, read, update, delete make up 90% of your daily work with any database.

find alone isn't enough — query operators let you fetch exactly the docs and fields you need.

Modeling in MongoDB is the opposite of SQL — design follows how you read the data, not the reverse.

Flexibility cuts both ways — without validation rules you can end up with messy, unexpected data.

The aggregation framework is MongoDB's analytics engine — any complex report is built as a pipeline.

Without an index MongoDB scans every doc — indexes are the line between a fast query and a server killer.

Sometimes several docs must change as one unit — transactions guarantee all-or-nothing.

One node = a single point of failure. Replica sets ensure availability and automatic recovery.

When data or load outgrows one server, sharding distributes them across many nodes.

In production, one bad query can slow the whole system — you must diagnose and tune with confidence.

MongoDB is more than storage — you can listen to changes in real time and build full-text search inside it.

An open, unauthenticated database is the #1 cause of data leaks — and production needs backups and monitoring.