MongoDB Reference

insertOne() inserts a single document and returns the _id of the inserted document. insertMany() accepts an array of documents and inserts them all in a single operation, returning an array of inserted IDs. For bulk imports, insertMany() is significantly faster than calling insertOne() in a loop because it batches the writes into fewer network round trips.

The aggregation pipeline is a sequence of stages where each stage transforms the input documents and passes the result to the next stage. Common stages include $match (filter documents like WHERE in SQL), $group (aggregate values like GROUP BY), $sort (order results), $project (reshape fields like SELECT), $lookup (join with another collection), and $unwind (flatten arrays into separate documents).

$lookup performs a left outer join between two collections. For example, joining orders with users: { $lookup: { from: "users", localField: "userId", foreignField: "_id", as: "user" } }. Use $lookup when your data uses the reference pattern (storing ObjectId references) and you need to combine data from multiple collections in a single query result.

A TTL (Time-To-Live) index automatically deletes documents after a specified number of seconds. Create one on a Date field: db.sessions.createIndex({ createdAt: 1 }, { expireAfterSeconds: 3600 }). This deletes session documents one hour after their createdAt timestamp. TTL indexes are commonly used for session management, cache entries, and log rotation.

The embedded document pattern stores related data inside the same document (e.g., address nested inside a user document). This makes reads fast since everything is retrieved in one query. The reference pattern stores a separate document and links them with an ObjectId, similar to a foreign key. Use embedding for one-to-one or one-to-few relationships that are always read together; use references for one-to-many relationships or when data is shared across multiple parent documents.

Run rs.initiate() with a configuration object specifying the set name and member hosts: rs.initiate({ _id: "myReplicaSet", members: [{ _id: 0, host: "mongo1:27017" }, { _id: 1, host: "mongo2:27017" }, { _id: 2, host: "mongo3:27017" }] }). MongoDB elects one member as primary and the others as secondaries. Check status with rs.status(). Applications connect using a connection string listing all replica set members.

Range sharding distributes documents across shards based on value ranges of the shard key — for example, timestamps so that recent data lands on specific shards. This is efficient for range queries but can create hotspots if data is written sequentially. Hash sharding hashes the shard key value to distribute documents more uniformly across shards, preventing write hotspots. Use hash sharding for randomly distributed keys like user IDs.

Run db.collection.find(query).explain("executionStats") to see how MongoDB executes the query. Look at the "executionStats" section: "totalKeysExamined" shows how many index keys were scanned, "totalDocsExamined" shows how many documents were loaded, and "nReturned" shows how many were returned. If totalDocsExamined is much larger than nReturned, the query lacks an efficient index. The "winningPlan" section shows whether an index scan (IXSCAN) or full collection scan (COLLSCAN) was used.