54 MongoDB interview questions with answers

All MongoDB interview questions

• What is MongoDB and when should you choose it?

• What are documents, collections and databases in MongoDB?

• How is BSON different from JSON?

• How do `findOne`, `find`, `insertOne`, and `updateOne` work in MongoDB?

• Should you embed or reference documents in MongoDB?

• How does MongoDB document size limit affect schema design?

• What do indexes help with in MongoDB and what are the trade-offs?

• How should you choose field order in a MongoDB compound index?

• What is a multikey index?

• When should you use a TTL index in MongoDB?

• What is `ObjectId` in MongoDB?

• What is `$lookup` in MongoDB and what are its risks?

• What is the purpose of `$unwind` in the aggregation pipeline?

• Does MongoDB support ACID transactions?

• What problem does sharding in MongoDB solve?

• What is an upsert in MongoDB?

• What is write concern in MongoDB?

• What is read preference in MongoDB?

• What is a `Cursor` in MongoDB and why is pagination needed?

• When should you use bulk operations in MongoDB?

• What is the oplog in MongoDB?

• What is the schema versioning pattern?

• What is a sharding key in MongoDB and how to choose a good one?

• How do you perform schema migrations in MongoDB?

• What is the difference between PostgreSQL B-tree index and MongoDB B-tree index?

• Why does MongoDB use BSON instead of JSON?

• When does the "16MB document size limit" error usually occur and how do you fix it?

• What are Capped Collections in MongoDB?

• How do the `$in` and `$or` operators differ in terms of performance?

• Explain "Eventual Consistency" when reading from a Secondary node.

• What is GridFS and when should you use it?

• How do multi-document transactions work in MongoDB?

• What is the difference between `updateOne` and `replaceOne`?

• How does locking work in MongoDB?

• How do you implement optimistic concurrency in MongoDB?

• How should multi-tenant MongoDB be designed?

• Where do you start debugging slow MongoDB queries?

• When should you not use MongoDB?

• After running `explain()`, what should you look at first to spot a problem?

• In an aggregation pipeline, why should `$match` come as early as possible?

• Why should you avoid letting an array inside a document grow without bound?

• What is a MongoDB change stream and what is it used for?

• Posts and comments: embed comments in the post or use a separate collection? What drives the decision?

• What is the ESR rule for ordering fields in a compound index? Why do sorted queries often hit a memory limit error?

• The index exists but the query still scans the whole collection — when does MongoDB skip an index?

• What is a covered query and how do you make one?

• How do partial and sparse indexes differ, and when do you use each?

• An aggregation fails with a 100MB limit at `$group`. Is enabling `allowDiskUse` the right fix?

• MongoDB is schemaless — how do you stop malformed data from entering a collection?

• A system writes one sensor reading per second per device. How do you design the collection to avoid billions of tiny documents?

• What actually goes wrong if you pick the wrong shard key, and can it be fixed?

• How do chunks and the balancer work in a sharded cluster? How do you deal with a jumbo chunk?

• The replica set primary dies mid-write — what happens to the in-flight writes?

• New code needs a field that 50 million existing documents lack. How do you roll it out with no downtime?