> For the complete MongoDB documentation index, see www.mongodb.com/docs/llms.txt

# Specify Validation Level for Existing Documents

For documents that already exist in your collection before you add validation, you can specify how MongoDB applies validation rules to these documents.

## Context

Your schema's `validationLevel` determines the documents for which MongoDB applies validation rules:

| Validation Level | Behavior |
| --- | --- |
| `moderate` | MongoDB applies the same validation rules to document inserts and updates to existing valid documents that match the validation rules. Updates to existing documents in the collection that don't match the validation rules aren't required to pass validation. |
| `strict` | (*Default*) MongoDB applies the same validation rules to all document inserts and updates. |
| `constraint` | MongoDB applies the validation rules to all document inserts and updates, and guarantees that every document in the collection satisfies the rules. Unlike `strict`, a collection with `constraint` validation can't contain documents that were invalid before you set the level. This level adds the following restrictions: The [validation action](https://www.mongodb.com/docs/manual/core/schema-validation/handle-invalid-documents.md#std-label-schema-validation-validationAction) must be `error` or `errorAndLog`.; You can't modify the schema validation rules while the level is `constraint`. To change your validation rules, downgrade to `strict`, modify the rules, and then upgrade back to `constraint`.; You cannot bypass document validation. For more information, see [Bypass Schema Validation.](https://www.mongodb.com/docs/manual/core/schema-validation/bypass-document-validation.md#std-label-schema-bypass-document-validation) **New in version 9.0** |

## Prerequisite

The examples on this page use a `movies` collection. To create the collection and insert the documents that the examples use, run the following command:

```javascript
db.movies.insertMany( [
   { title: "The Godfather", year: 1972 },
   { title: "Casablanca", year: 1942 }
] )

```

## Specify moderate Validation

The following example adds a `moderate` validator to the `movies` collection and inserts a document that satisfies the validation rules.

1. Specify validation rules with `moderate` validation level.

   Add a validator to the `movies` collection with `moderate` `validationLevel`:

   ```javascript
   db.runCommand( {
      collMod: "movies",
      validator: { $jsonSchema: {
         bsonType: "object",
         required: [ "title" ],
         properties: {
            title: {
               bsonType: "string",
               description: "title must be a string and is required"
            }
         }
      } },
      validationLevel: "moderate"
   } )

   ```

   Because the `validationLevel` is `moderate`, MongoDB applies the validation rules to all document inserts and to updates of existing documents that already meet the rules.

2. Insert a valid document.

   The following insert command adds a document that includes a `title` field with a string value, which satisfies the validation rule:

   ```javascript
   db.movies.insertOne(
      { title: "Wings", year: 1927 }
   )

   ```

3. Observe results.

   The insert succeeds because the document satisfies the validation rules. MongoDB returns output similar to the following:

   ```javascript
   {
     acknowledged: true,
     insertedId: ...
   }

   ```

## Specify strict Validation

The following example adds a `strict` validator to the `movies` collection and inserts a document that satisfies the validation rules.

1. Specify validation rules with `strict` validation level.

   Add a validator to the `movies` collection with `strict` `validationLevel`:

   ```javascript
   db.runCommand( {
      collMod: "movies",
      validator: { $jsonSchema: {
         bsonType: "object",
         required: [ "title" ],
         properties: {
            title: {
               bsonType: "string",
               description: "title must be a string and is required"
            }
         }
      } },
      validationLevel: "strict"
   } )

   ```

   Because the `validationLevel` is `strict`, MongoDB applies the validation rules to all document inserts and updates.

2. Insert a valid document.

   The following insert command adds a document that includes a `title` field with a string value, which satisfies the validation rule:

   ```javascript
   db.movies.insertOne(
      { title: "The Conversation", year: 1974 }
   )

   ```

3. Observe results.

   The insert succeeds because the document satisfies the validation rules. MongoDB returns output similar to the following:

   ```javascript
   {
     acknowledged: true,
     insertedId: ...
   }

   ```

## Specify constraint Validation

The following example upgrades the `movies` collection to the `constraint` validation level. When you upgrade to the `constraint` level, MongoDB checks that every document already in the collection satisfies the validator.

**Note:**

The following procedure applies to a collection that already exists. If you set the `validationLevel` to `constraint` when you create the collection, you don't need to set the level to `strict` first or prepare the collection.

1. (Conditional) Set your `validationLevel` to `strict`.

   If your collection's `validationLevel` is not already `strict`, follow the procedure to [specify strict validation.](https://www.mongodb.com/docs/manual/core/schema-validation/specify-validation-level.md#std-label-schema-specify-validation-level-strict)

2. Prepare the collection for the `constraint` level.

   Run [`collMod`](https://www.mongodb.com/docs/manual/reference/command/collmod.md#mongodb-dbcommand-dbcmd.collMod) with `prepareConstraintValidationLevel` set to `true`:

   ```javascript
   db.runCommand( {
      collMod: "movies",
      prepareConstraintValidationLevel: true
   } )

   ```

   This command prepares the collection for the upgrade by blocking `bypassDocumentValidation` and preventing changes to the validator. To learn how MongoDB restricts bypassing, see [Bypass Schema Validation.](https://www.mongodb.com/docs/manual/core/schema-validation/bypass-document-validation.md#std-label-schema-bypass-document-validation)

3. Set the validation level to `constraint`.

   Run [`collMod`](https://www.mongodb.com/docs/manual/reference/command/collmod.md#mongodb-dbcommand-dbcmd.collMod) to set the `validationLevel` to `constraint`. The `validationAction` must be `error` or `errorAndLog`:

   ```javascript
   db.runCommand( {
      collMod: "movies",
      validationLevel: "constraint",
      validationAction: "error"
   } )

   ```

   You can't change the `validator` in the same `collMod` operation that upgrades the `validationLevel` to `constraint`. Set the validation rules before you run this command. To change the rules later, downgrade the level to `strict`, modify the rules, and then upgrade back to `constraint`.

   Because the `validationLevel` is `constraint`, MongoDB applies the validation rules to all document inserts and updates, regardless of whether an existing document met the rules before the upgrade.

   When you set the level to `constraint`, MongoDB scans all existing documents in the collection and fails the `collMod` operation if any document violates the validator. This scan can be computationally expensive for large collections. If the operation fails, update the violating documents to pass the validator before you continue.

4. Insert a valid document.

   The following insert command adds a document that includes a `title` field with a string value, which satisfies the validation rule:

   ```javascript
   db.movies.insertOne(
      { title: "Sunrise", year: 1927 }
   )

   ```

5. Observe results.

   The insert succeeds because the document satisfies the validation rules. MongoDB returns output similar to the following:

   ```javascript
   {
     acknowledged: true,
     insertedId: ...
   }

   ```

**Important:**

The error output is intended for human consumption. It may change in the future and should not be relied upon in scripts.

## Learn More

- [Choose How to Handle Invalid Documents](https://www.mongodb.com/docs/manual/core/schema-validation/handle-invalid-documents.md#std-label-schema-validation-handle-invalid-docs)

- [Modify Schema Validation](https://www.mongodb.com/docs/manual/core/schema-validation/update-schema-validation.md#std-label-schema-update-validation)