Schema design, $lookup, compound indexes, change streams.

MongoDB — Documents, Aggregation Pipeline, and Indexing

MongoDB stores data as BSON documents, enabling flexible schemas and nested structures. The aggregation pipeline ($match, $group, $lookup, $project) replaces SQL joins and GROUP BY. Indexes use B-trees; compound indexes follow the ESR rule (Equality, Sort, Range). Change streams enable real-time reactions to data changes without polling. Multi-document ACID transactions are available since v4.0 but add overhead — prefer single-document operations where possible.

Interactive Pipeline Flow

UnderstandImplementOptimise

🔌 Connect — replica set URI

Connect to a replica set URI: mongodb://host1,host2,host3/?replicaSet=rs0. Atlas provides managed replica sets. Use connection pooling (default pool size: 100).

Architecture Diagrams

Aggregation pipeline for total revenue per customer.

Loading diagram...

Key Design Decisions

📄Design for your queries, not a generic schema

MongoDB schema design is query-driven. Embed data that is always read together. Reference data that is queried independently or grows without bound.

📑ESR rule for compound indexes

Equality → Sort → Range. If your query is WHERE userId=X ORDER BY createdAt, the index should be {userId:1, createdAt:-1}. Putting range before sort wastes the sort optimization.

🔍Always explain() before deploying queries

db.collection.find(query).explain("executionStats") shows whether a COLLSCAN (full scan) or IXSCAN is used. A COLLSCAN on a large collection is an outage waiting to happen.

Discussion