Elastic MongoDB connector reference

edit

Elastic MongoDB connector reference

edit

The Elastic MongoDB connector is a connector for MongoDB data sources. This connector is written in Python using the Elastic connector framework.

View the source code for this connector (branch 8.14, compatible with Elastic 8.14).

Native connector (Elastic Cloud)

edit
View native connector reference

Availability and prerequisites

edit

This connector is available as a native connector in Elastic versions 8.5.0 and later.

To use this connector natively in Elastic Cloud, satisfy all native connector requirements.

Compatibility

edit

This connector is compatible with MongoDB Atlas and MongoDB 3.6 and later.

The data source and your Elastic deployment must be able to communicate with each other over a network.

Configuration

edit

Each time you create an index to be managed by this connector, you will create a new connector configuration. You will need some or all of the following information about the data source.

Server hostname

The URI of the MongoDB host. Examples:

  • mongodb+srv://my_username:my_password@cluster0.mongodb.net/mydb?w=majority
  • mongodb://127.0.0.1:27017
Username

The MongoDB username the connector will use.

The user must have access to the configured database and collection. You may want to create a dedicated, read-only user for each connector.

Password
The MongoDB password the connector will use.
Database
The MongoDB database to sync. The database must be accessible using the configured username and password.
Collection
The MongoDB collection to sync. The collection must exist within the configured database. The collection must be accessible using the configured username and password.
Direct connection
Toggle to use the direct connection option for the MongoDB client. Disabled by default.
SSL/TLS Connection

Toggle to establish a secure connection to the MongoDB server using SSL/TLS encryption. Ensure that your MongoDB deployment supports SSL/TLS connections. Enable if your MongoDB cluster uses DNS SRV records (namely MongoDB Atlas users).

Disabled by default.

Certificate Authority (.pem)
Specifies the root certificate from the Certificate Authority. The value of the certificate is used to validate the certificate presented by the MongoDB instance.

Atlas users can leave this blank because Atlas uses a widely trusted root CA.

Skip certificate verification
Skips various certificate validations (if SSL is enabled). Disabled by default.

We strongly recommend leaving this option disabled in production environments.

Create a MongoDB connector

edit
Use the UI
edit

To create a new MongoDB connector:

  1. Navigate to the Search → Connectors page in the Kibana UI.
  2. Follow the instructions to create a new native MongoDB connector.

For additional operations, see Using connectors.

Use the API
edit

You can use the Elasticsearch Create connector API to create a new native MongoDB connector.

For example:

PUT _connector/<my-mongodb-connector>
{
  "index_name": "<my-elasticsearch-index>",
  "name": "Content synced from MongoDB",
  "service_type": "mongodb",
  "is_native": "true"
}
You’ll also need to create an API key for the connector to use.

The user needs the cluster privileges manage_api_key and write_connector_secrets to generate API keys programmatically.

To create an API key for the connector:

  1. Run the following command, replacing values where indicated. Note the id and encoded return values from the response:

    POST /_security/api_key
    {
      "name": "<connector_name>-connector-api-key",
      "role_descriptors": {
        "<connector_name>-connector-role": {
          "cluster": [
            "monitor"
          ],
          "indices": [
            {
              "names": [
                "<index_name>",
                ".search-acl-filter-<index_name>",
                ".elastic-connectors*"
              ],
              "privileges": [
                "all"
              ],
              "allow_restricted_indices": false
            }
          ]
        }
      }
    }
  2. Use the encoded value to store a connector secret, and note the id return value from this response:

    POST _connector/_secret
    {
      "value": <encoded_api_key>
    }
  3. Use the API key id and the connector secret id to update the connector:

    PUT /_connector/<connector_id>/_api_key_id
    {
      "api_key_id": "<API key id>",
      "api_key_secret_id": "<secret id>"
    }

Refer to the Elasticsearch API documentation for details of all available Connector APIs.

Usage

edit

To use this connector as a native connector, use the Connector workflow. See Native connectors (managed service).

For additional operations, see Using connectors.

Example

edit

An example is available for this connector. See Native connector tutorial (MongoDB).

Known issues

edit
SSL must be enabled for MongoDB Atlas
edit
  • A bug introduced in 8.12.0 causes the connector to fail to sync Mongo Atlas urls (mongo+srv) unless SSL/TLS is enabled.
Expressions and variables in aggregation pipelines
edit

It’s not possible to use expressions like new Date() inside an aggregation pipeline. These expressions won’t be evaluated by the underlying MongoDB client, but will be passed as a string to the MongoDB instance. A possible workaround is to use aggregation variables.

Incorrect (new Date() will be interpreted as string):

{
    "aggregate": {
        "pipeline": [
            {
                "$match": {
                  "expiresAt": {
                    "$gte": "new Date()"
                  }
                }
            }
        ]
    }
}

Correct (usage of $$NOW):

{
  "aggregate": {
    "pipeline": [
      {
        "$addFields": {
          "current_date": {
            "$toDate": "$$NOW"
          }
        }
      },
      {
        "$match": {
          "$expr": {
            "$gte": [
              "$expiresAt",
              "$current_date"
            ]
          }
        }
      }
    ]
  }
}
Connecting with self-signed or custom CA TLS Cert
edit

Currently, the MongoDB connector does not support working with self-signed or custom CA certs when connecting to your self-managed MongoDB host.

The following workaround should not be used in production.

This can be worked around in development environments, by appending certain query parameters to the configured host.

For example, if your host is mongodb+srv://my.mongo.host.com, appending ?tls=true&tlsAllowInvalidCertificates=true will allow disabling TLS certificate verification.

The full host in this example will look like this:

mongodb+srv://my.mongo.host.com/?tls=true&tlsAllowInvalidCertificates=true

See Known issues for any issues affecting all connectors.

Troubleshooting

edit

See Troubleshooting.

Security

edit

See Security.

Documents and syncs

edit

The following describes the default syncing behavior for this connector. Use sync rules and ingest pipelines to customize syncing for specific indices.

All documents in the configured MongoDB database and collection are extracted and transformed into documents in your Elasticsearch index.

  • The connector creates one Elasticsearch document for each MongoDB document in the configured database and collection.
  • For each document, the connector transforms each MongoDB field into an Elasticsearch field.
  • For each field, Elasticsearch dynamically determines the data type.

This results in Elasticsearch documents that closely match the original MongoDB documents.

The Elasticsearch mapping is created when the first document is created.

Each sync is a "full" sync. For each MongoDB document discovered:

  • If it does not exist, the document is created in Elasticsearch.
  • If it already exists in Elasticsearch, the Elasticsearch document is replaced and the version is incremented.
  • If an existing Elasticsearch document no longer exists in the MongoDB collection, it is deleted from Elasticsearch.
  • Embedded documents are stored as an object field in the parent document.

This is recursive, because embedded documents can themselves contain embedded documents.

  • Files bigger than 10 MB won’t be extracted
  • Permissions are not synced. All documents indexed to an Elastic deployment will be visible to all users with access to that Elastic Deployment.

Sync rules

edit

The following sections describe Sync rules for this connector.

Basic sync rules are identical for all connectors and are available by default.

Advanced rules for MongoDB can be used to express either find queries or aggregation pipelines. They can also be used to tune options available when issuing these queries/pipelines.

find queries
edit

You must create a text index on the MongoDB collection in order to perform text searches.

For find queries, the structure of this JSON DSL should look like:

{
	"find":{
		"filter": {
			// find query goes here
		},
		"options":{
			// query options go here
		}
	}
}

For example:

{
	"find": {
		"filter": {
			"$text": {
				"$search": "garden",
				"$caseSensitive": false
			}
		},
		"skip": 10,
		"limit": 1000
	}
}

find queries also support additional options, for example the projection object:

{
  "find": {
    "filter": {
      "languages": [
        "English"
      ],
      "runtime": {
        "$gt":90
      }
    },
    "projection":{
      "tomatoes": 1
    }
  }
}

Where the available options are:

  • allow_disk_use (true, false) — When set to true, the server can write temporary data to disk while executing the find operation. This option is only available on MongoDB server versions 4.4 and newer.
  • allow_partial_results (true, false) — Allows the query to get partial results if some shards are down.
  • batch_size (Integer) — The number of documents returned in each batch of results from MongoDB.
  • filter (Object) — The filter criteria for the query.
  • limit (Integer) — The max number of docs to return from the query.
  • max_time_ms (Integer) — The maximum amount of time to allow the query to run, in milliseconds.
  • no_cursor_timeout (true, false) — The server normally times out idle cursors after an inactivity period (10 minutes) to prevent excess memory use. Set this option to prevent that.
  • projection (Array, Object) — The fields to include or exclude from each doc in the result set. If an array, it should have at least one item.
  • return_key (true, false) — Return index keys rather than the documents.
  • show_record_id (true, false) — Return the $recordId for each doc in the result set.
  • skip (Integer) — The number of docs to skip before returning results.
Aggregation pipelines
edit

Similarly, for aggregation pipelines, the structure of the JSON DSL should look like:

{
	"aggregate":{
		"pipeline": [
			// pipeline elements go here
		],
		"options": {
            // pipeline options go here
		}
    }
}

Where the available options are:

  • allowDiskUse (true, false) — Set to true if disk usage is allowed during the aggregation.
  • batchSize (Integer) — The number of documents to return per batch.
  • bypassDocumentValidation (true, false) — Whether or not to skip document level validation.
  • collation (Object) — The collation to use.
  • comment (String) — A user-provided comment to attach to this command.
  • hint (String) — The index to use for the aggregation.
  • let (Object) — Mapping of variables to use in the pipeline. See the server documentation for details.
  • maxTimeMs (Integer) — The maximum amount of time in milliseconds to allow the aggregation to run.

Migrating from the Ruby connector framework

edit

As part of the 8.8.0 release the MongoDB connector was moved from the Ruby connectors framework to the Elastic connector framework.

This change introduces minor formatting modifications to data ingested from MongoDB:

  1. Nested object id field name has changed from "_id" to "id". For example, if you had a field "customer._id", this will now be named "customer.id".
  2. Date format has changed from YYYY-MM-DD'T'HH:mm:ss.fff'Z' to YYYY-MM-DD'T'HH:mm:ss

If your MongoDB connector stopped working after migrating from 8.7.x to 8.8.x, read the workaround outlined in Known issues. If that does not work, we recommend deleting the search index attached to this connector and re-creating a MongoDB connector from scratch.

Connector client (self-managed)

edit
View connector client reference

Availability and prerequisites

edit

This connector is also available as a connector client from the Elastic connector framework. To use this connector as a connector client, satisfy all connector client requirements.

Compatibility

edit

This connector is compatible with MongoDB Atlas and MongoDB 3.6 and later.

The data source and your Elastic deployment must be able to communicate with each other over a network.

Configuration

edit

When using the connector client workflow, initially these fields will use the default configuration set in the connector source code. These are set in the get_default_configuration function definition.

These configurable fields will be rendered with their respective labels in the Kibana UI. Once connected, you’ll be able to update these values in Kibana.

The following configuration fields are required to set up the connector:

host

The URI of the MongoDB host. Examples:

  • mongodb+srv://my_username:my_password@cluster0.mongodb.net/mydb?w=majority
  • mongodb://127.0.0.1:27017
user

The MongoDB username the connector will use.

The user must have access to the configured database and collection. You may want to create a dedicated, read-only user for each connector.

password
The MongoDB password the connector will use.

Anonymous authentication is supported for testing purposes only, but should not be used in production. Omit the username and password, to use default values.

database
The MongoDB database to sync. The database must be accessible using the configured username and password.
collection
The MongoDB collection to sync. The collection must exist within the configured database. The collection must be accessible using the configured username and password.
direct_connection
Whether to use the direct connection option for the MongoDB client. Default value is False.
ssl_enabled

Whether to establish a secure connection to the MongoDB server using SSL/TLS encryption. Ensure that your MongoDB deployment supports SSL/TLS connections. Enable if your MongoDB cluster uses DNS SRV records (namely MongoDB Atlas users).

Default value is False.

ssl_ca
Specifies the root certificate from the Certificate Authority. The value of the certificate is used to validate the certificate presented by the MongoDB instance.

Atlas users can leave this blank because Atlas uses a widely trusted root CA.

tls_insecure
Skips various certificate validations (if SSL is enabled). Default value is False.

We strongly recommend leaving this option disabled in production environments.

Create a MongoDB connector

edit
Use the UI
edit

To create a new MongoDB connector:

  1. Navigate to the Search → Connectors page in the Kibana UI.
  2. Follow the instructions to create a new MongoDB connector client.

For additional operations, see Using connectors.

Use the API
edit

You can use the Elasticsearch Create connector API to create a new self-managed MongoDB connector client.

For example:

PUT _connector/my-mongodb-connector
{
  "index_name": "my-elasticsearch-index",
  "name": "Content synced from MongoDB",
  "service_type": "mongodb"
}
You’ll also need to create an API key for the connector to use.

The user needs the cluster privileges manage_api_key and write_connector_secrets to generate API keys programmatically.

To create an API key for the connector:

  1. Run the following command, replacing values where indicated. Note the encoded return values from the response:

    POST /_security/api_key
    {
      "name": "<connector_name>-connector-api-key",
      "role_descriptors": {
        "<connector_name>-connector-role": {
          "cluster": [
            "monitor"
          ],
          "indices": [
            {
              "names": [
                "<index_name>",
                ".search-acl-filter-<index_name>",
                ".elastic-connectors*"
              ],
              "privileges": [
                "all"
              ],
              "allow_restricted_indices": false
            }
          ]
        }
      }
    }
  2. Update your config.yml file with the API key encoded value.

Refer to the Elasticsearch API documentation for details of all available Connector APIs.

Usage

edit

To use this connector as a connector client, see Connector clients (self-managed) For additional usage operations, see Using connectors.

Example

edit

An example is available for this connector. See Native connector tutorial (MongoDB).

Known issues

edit
SSL must be enabled for MongoDB Atlas
edit
  • A bug introduced in 8.12.0 causes the connector to fail to sync Mongo Atlas urls (mongo+srv) unless SSL/TLS is enabled.
Expressions and variables in aggregation pipelines
edit

It’s not possible to use expressions like new Date() inside an aggregation pipeline. These expressions won’t be evaluated by the underlying MongoDB client, but will be passed as a string to the MongoDB instance. A possible workaround is to use aggregation variables.

Incorrect (new Date() will be interpreted as string):

{
    "aggregate": {
        "pipeline": [
            {
                "$match": {
                  "expiresAt": {
                    "$gte": "new Date()"
                  }
                }
            }
        ]
    }
}

Correct (usage of $$NOW):

{
  "aggregate": {
    "pipeline": [
      {
        "$addFields": {
          "current_date": {
            "$toDate": "$$NOW"
          }
        }
      },
      {
        "$match": {
          "$expr": {
            "$gte": [
              "$expiresAt",
              "$current_date"
            ]
          }
        }
      }
    ]
  }
}
Connecting with self-signed or custom CA TLS Cert
edit

Currently, the MongoDB connector does not support working with self-signed or custom CA certs when connecting to your self-managed MongoDB host.

The following workaround should not be used in production.

This can be worked around in development environments, by appending certain query parameters to the configured host.

For example, if your host is mongodb+srv://my.mongo.host.com, appending ?tls=true&tlsAllowInvalidCertificates=true will allow disabling TLS certificate verification.

The full host in this example will look like this:

mongodb+srv://my.mongo.host.com/?tls=true&tlsAllowInvalidCertificates=true

Docker image errors out for versions 8.12.0 and 8.12.1
edit

A bug introduced in 8.12.0 causes the Connectors docker image to error out if run using MongoDB as its source. The command line will output the error cannot import name 'coroutine' from 'asyncio'. This issue is fixed in versions 8.12.2 and 8.13.0. This bug does not affect native connectors.

See Known issues for any issues affecting all connectors.

Troubleshooting

edit

See Troubleshooting.

Security

edit

See Security.

Deployment using Docker

edit

You can deploy the MongoDB connector as a self-managed connector client using Docker. Follow these instructions.

Step 1: Download sample configuration file

Download the sample configuration file. You can either download it manually or run the following command:

curl https://raw.githubusercontent.com/elastic/connectors/main/config.yml.example --output ~/connectors-config/config.yml

Remember to update the --output argument value if your directory name is different, or you want to use a different config file name.

Step 2: Update the configuration file for your self-managed connector

Update the configuration file with the following settings to match your environment:

  • elasticsearch.host
  • elasticsearch.api_key
  • connectors

If you’re running the connector service against a Dockerized version of Elasticsearch and Kibana, your config file will look like this:

# When connecting to your cloud deployment you should edit the host value
elasticsearch.host: http://host.docker.internal:9200
elasticsearch.api_key: <ELASTICSEARCH_API_KEY>

connectors:
  -
    connector_id: <CONNECTOR_ID_FROM_KIBANA>
    service_type: mongodb
    api_key: <CONNECTOR_API_KEY_FROM_KIBANA> # Optional. If not provided, the connector will use the elasticsearch.api_key instead

Using the elasticsearch.api_key is the recommended authentication method. However, you can also use elasticsearch.username and elasticsearch.password to authenticate with your Elasticsearch instance.

Note: You can change other default configurations by simply uncommenting specific settings in the configuration file and modifying their values.

Step 3: Run the Docker image

Run the Docker image with the Connector Service using the following command:

docker run \
-v ~/connectors-config:/config \
--network "elastic" \
--tty \
--rm \
docker.elastic.co/enterprise-search/elastic-connectors:8.14.3.0 \
/app/bin/elastic-ingest \
-c /config/config.yml

Refer to DOCKER.md in the elastic/connectors repo for more details.

Find all available Docker images in the official registry.

We also have a quickstart self-managed option using Docker Compose, so you can spin up all required services at once: Elasticsearch, Kibana, and the connectors service. Refer to this README in the elastic/connectors repo for more information.

Documents and syncs

edit

The following describes the default syncing behavior for this connector. Use sync rules and ingest pipelines to customize syncing for specific indices.

All documents in the configured MongoDB database and collection are extracted and transformed into documents in your Elasticsearch index.

  • The connector creates one Elasticsearch document for each MongoDB document in the configured database and collection.
  • For each document, the connector transforms each MongoDB field into an Elasticsearch field.
  • For each field, Elasticsearch dynamically determines the data type.

This results in Elasticsearch documents that closely match the original MongoDB documents.

The Elasticsearch mapping is created when the first document is created.

Each sync is a "full" sync. For each MongoDB document discovered:

  • If it does not exist, the document is created in Elasticsearch.
  • If it already exists in Elasticsearch, the Elasticsearch document is replaced and the version is incremented.
  • If an existing Elasticsearch document no longer exists in the MongoDB collection, it is deleted from Elasticsearch.
  • Embedded documents are stored as an object field in the parent document.

This is recursive, because embedded documents can themselves contain embedded documents.

  • Files bigger than 10 MB won’t be extracted
  • Permissions are not synced. All documents indexed to an Elastic deployment will be visible to all users with access to that Elastic Deployment.

Sync rules

edit

The following sections describe Sync rules for this connector.

Basic sync rules are identical for all connectors and are available by default.

Advanced rules for MongoDB can be used to express either find queries or aggregation pipelines. They can also be used to tune options available when issuing these queries/pipelines.

find queries
edit

You must create a text index on the MongoDB collection in order to perform text searches.

For find queries, the structure of this JSON DSL should look like:

{
	"find":{
		"filter": {
			// find query goes here
		},
		"options":{
			// query options go here
		}
	}
}

For example:

{
	"find": {
		"filter": {
			"$text": {
				"$search": "garden",
				"$caseSensitive": false
			}
		},
		"skip": 10,
		"limit": 1000
	}
}

find queries also support additional options, for example the projection object:

{
  "find": {
    "filter": {
      "languages": [
        "English"
      ],
      "runtime": {
        "$gt":90
      }
    },
    "projection":{
      "tomatoes": 1
    }
  }
}

Where the available options are:

  • allow_disk_use (true, false) — When set to true, the server can write temporary data to disk while executing the find operation. This option is only available on MongoDB server versions 4.4 and newer.
  • allow_partial_results (true, false) — Allows the query to get partial results if some shards are down.
  • batch_size (Integer) — The number of documents returned in each batch of results from MongoDB.
  • filter (Object) — The filter criteria for the query.
  • limit (Integer) — The max number of docs to return from the query.
  • max_time_ms (Integer) — The maximum amount of time to allow the query to run, in milliseconds.
  • no_cursor_timeout (true, false) — The server normally times out idle cursors after an inactivity period (10 minutes) to prevent excess memory use. Set this option to prevent that.
  • projection (Array, Object) — The fields to include or exclude from each doc in the result set. If an array, it should have at least one item.
  • return_key (true, false) — Return index keys rather than the documents.
  • show_record_id (true, false) — Return the $recordId for each doc in the result set.
  • skip (Integer) — The number of docs to skip before returning results.
Aggregation pipelines
edit

Similarly, for aggregation pipelines, the structure of the JSON DSL should look like:

{
	"aggregate":{
		"pipeline": [
			// pipeline elements go here
		],
		"options": {
            // pipeline options go here
		}
    }
}

Where the available options are:

  • allowDiskUse (true, false) — Set to true if disk usage is allowed during the aggregation.
  • batchSize (Integer) — The number of documents to return per batch.
  • bypassDocumentValidation (true, false) — Whether or not to skip document level validation.
  • collation (Object) — The collation to use.
  • comment (String) — A user-provided comment to attach to this command.
  • hint (String) — The index to use for the aggregation.
  • let (Object) — Mapping of variables to use in the pipeline. See the server documentation for details.
  • maxTimeMs (Integer) — The maximum amount of time in milliseconds to allow the aggregation to run.

Migrating from the Ruby connector framework

edit

As part of the 8.8.0 release the MongoDB connector was moved from the Ruby connectors framework to the Elastic connector framework.

This change introduces minor formatting modifications to data ingested from MongoDB:

  1. Nested object id field name has changed from "_id" to "id". For example, if you had a field "customer._id", this will now be named "customer.id".
  2. Date format has changed from YYYY-MM-DD'T'HH:mm:ss.fff'Z' to YYYY-MM-DD'T'HH:mm:ss

If your MongoDB connector stopped working after migrating from 8.7.x to 8.8.x, read the workaround outlined in Known issues. If that does not work, we recommend deleting the search index attached to this connector and re-creating a MongoDB connector from scratch.