MongoDB

MongoDB interviews test document modeling for access patterns: embedding versus referencing, index design for compound queries, aggregation pipelines, and when multi-document transactions are worth the cost.

Subtopics: documents, collections, indexes, aggregation pipeline, schema design, transactions.

On interviews: sketch a document shape for a read-heavy feature, name indexes for the pipeline, and explain growth and write contention risks.

Common pitfalls: relational modeling with extra joins in app code; unbounded document growth; transactions as default instead of document boundaries.

The trade-off is balancing schema flexibility against index discipline and operational complexity at scale.

Checklist:

• Model around read locality and update boundaries.
• Index real query and sort shapes.
• Use aggregation for server-side reshape, not app fan-out.
• Reserve transactions for cross-document invariants only.