LearnAI ToolsCareerPractice BuildsPlayContact
Lesson 2316 min read

Schema Validation

Enforce structure and data types on a collection using MongoDB's JSON Schema validation rules.

Why Add Validation to a Flexible-Schema Database?

MongoDB's flexibility is a feature, not an excuse to skip data integrity entirely. Schema validation lets you enforce required fields, correct types, and value constraints at the database level — a safety net that catches malformed documents even if application-level validation has a bug.

Defining a Validator

A validator is defined using JSON Schema syntax, passed when creating a collection.

db.createCollection("users", {
  validator: {
    $jsonSchema: {
      bsonType: "object",
      required: ["name", "email"],
      properties: {
        name: { bsonType: "string", description: "must be a string and is required" },
        email: { bsonType: "string", pattern: "^.+@.+$", description: "must be a valid email" },
        age: { bsonType: "int", minimum: 0, maximum: 150 },
      },
    },
  },
});

Attempting to insert a document that violates this schema is rejected with a clear validation error.

Validation Levels

The validationLevel and validationAction options control how strictly and how loudly validation is enforced.

OptionValuesMeaning
validationLevelstrict (default) / moderatestrict validates all inserts and updates; moderate only validates documents that already passed validation before
validationActionerror (default) / warnerror rejects invalid writes; warn logs a warning but allows the write through
warn Is Useful for Migrating Existing Data

Setting validationAction: "warn" temporarily lets you introduce a new schema without breaking existing writes, while surfacing which documents don't yet conform — useful for a gradual migration.

Adding Validation to an Existing Collection

A validator can be added or changed on an already-existing collection with collMod, without needing to recreate it.

db.runCommand({
  collMod: "users",
  validator: { $jsonSchema: { /* ... */ } },
  validationLevel: "moderate",
});

FAQs

No — it's a complementary safety net at the database layer; application-level validation still matters for things like showing helpful error messages to a user before a request even reaches the database.

The overhead is generally small and well worth the data integrity guarantee it provides, though very complex schemas can add measurable cost — profile if you have concerns at extreme write volumes.

Summary

JSON Schema validation lets you enforce structure and types even in a flexible-schema database, with tunable strictness for migrations. With schema design covered, the next section moves into indexes — the key to making all of this fast at scale.

Next Lesson →

Introduction to Indexes