Docs Menu

Docs HomeMongoDB Kafka Connector

All Source Connector Configuration Properties

On this page

  • Overview
  • MongoDB Connection
  • Kafka Topic
  • Change Streams
  • Output Format
  • Startup
  • Error Handling and Resuming from Interruption

On this page, you can view all available configuration properties for your MongoDB Kafka source connector. This page duplicates the content of the other source connector configuration properties pages.

To view a list of all source connector configuration properties pages, see the Source Connector Configuration Properties page.

Use the following configuration settings to specify how your MongoDB Kafka source connector establishes a connection and communicates with your MongoDB cluster.

To view only the options related to your MongoDB connection, see the MongoDB Source Connection Properties page.

Name
Description
connection.uri
Required

Type: string

Description:
The URI connection string to connect to your MongoDB instance or cluster.
For more information, see the Connect to MongoDB guide

Important

Avoid Exposing Your Authentication Credentials

To avoid exposing your authentication credentials in your connection.uri setting, use a ConfigProvider and set the appropriate configuration parameters.

Default: mongodb://localhost:27017,localhost:27018,localhost:27019
Accepted Values: A MongoDB URI connection string
database
Type: string

Description:
Name of the database to watch for changes. If not set, the connector watches all databases for changes.

Default: ""
Accepted Values: A single database name
collection
Type: string

Description:
Name of the collection in the database to watch for changes. If not set, the connector watches all collections for changes.

Important

If your database configuration is set to "", the connector ignores the collection setting.


Default: ""
Accepted Values: A single collection name
server.api.version
Type: string

Description:
The Stable API version you want to use with your MongoDB cluster. For more information on the Stable API and versions of MongoDB Server that support it, see the Stable API guide.

Default: ""
Accepted Values: An empty string or a valid Stable API version.
server.api.deprecationErrors
Type: boolean

Description:
When set to true, if the connector calls a command on your MongoDB instance that's deprecated in the declared Stable API version, it raises an exception.
You can set the API version with the server.api.version configuration option. For more information on the Stable API, see the MongoDB manual entry on the Stable API.

Default: false
Accepted Values: true or false
server.api.strict
Type: boolean

Description:
When set to true, if the connector calls a command on your MongoDB instance that's not covered in the declared Stable API version, it raises an exception.
You can set the API version with the server.api.version configuration option. For more information on the Stable API, see the MongoDB manual entry on the Stable API.

Default: false
Accepted Values: true or false

Use the following configuration settings to specify which Kafka topics the MongoDB Kafka source connector should publish data to.

To view only the options related to your Kafka topic, see the Kafka Topic Properties page.

Name
Description
topic.prefix
Type: string

Description:
The prefix to prepend to database and collection names to generate the name of the Kafka topic on which to publish the data.

Tip

Default: ""
Accepted Values: A string composed of ASCII alphanumeric characters including ".", "-", and "_"
topic.suffix
Type: string

Description:
The suffix to append to database and collection names to generate the name of the Kafka topic on which to publish the data.
Default: ""
Accepted Values: A string composed of ASCII alphanumeric characters including ".", "-", and "_"
topic.namespace.map
Type: string

Description:
Specifies a JSON mapping between change stream document namespaces and topic names.

Example

The following mapping instructs the connector to perform the following actions:

  • Publish change stream documents originating from the myDb.myColl MongoDB collection to the topicOne Kafka topic.

  • Publish change stream documents originating from the myDb MongoDB database and a collection other than myColl to the Kafka topic topicTwo.<collectionName>.

  • Publish change stream documents originating from the myDb MongoDB database that do not bear a collection name to the topicTwo Kafka topic.

topic.namespace.map={"myDb.myColl": "topicOne", "myDb": "topicTwo"}

The following examples show which topic the connector sends different change stream documents to using the preceding mapping:

  • A change stream document originating from the myDb database and the myColl collection publishes to the topicOne topic.

  • A change stream document originating from the myDb database and the newColl collection publishes to the topicTwo.newColl topic.

  • A change stream document originating from dropping the myDb database, which does not bear a collection name, publishes to the topicTwo topic.

You can use the "*" wildcard character to match change stream document namespaces.

Example

The following mapping instructs the connector to publish all change stream documents to the topicThree topic:

topic.namespace.map={"*": "topicThree"}
Default: ""
Accepted Values: A valid JSON object
topic.separator
Type: string

Description:
Specifies the string the connector uses to concatenate the values used to create the name of your topic. The connector publishes records to a topic with a name formed by concatenating the values of the following fields in the following order:
  1. topic.prefix

  2. database

  3. collection

  4. topic.suffix

Example

The following configuration instructs the connector to publish change stream documents from the coll collection of the db database to the prefix-db-coll topic:

topic.prefix=prefix
database=db
collection=coll
topic.separator=-

Important

Topic Separator and Topic Namespace Map

When you use the topic.separator property, keep in mind that it does not affect how you define the topic.namespace.map property. The topic.namespace.map property uses MongoDB namespaces which you must always specify with a "." character to separate the database and collection name.


Default: "."
Accepted Values: A string
topic.mapper
Type: string

Description:
The Java class that defines your custom topic mapping logic.

Default: com.mongodb.kafka.connect.source.topic.mapping.DefaultTopicMapper
Accepted Values: Valid full class name of an implementation of the TopicMapper class.

Use the following configuration settings to specify aggregation pipelines for change streams and read preferences for change stream cursors when working with the MongoDB Kafka source connector.

To view only the options related to change streams, see the Change Stream Properties page.

Name
Description
pipeline
Type: string

Description:
An array of aggregation pipelines to run in your change stream. You must configure this setting for the change stream event document, not the fullDocument field.

Example

[{"$match": { "$and": [{"operationType": "insert"}, {"fullDocument.eventId": 1404 }] } }]
Default: []
Accepted Values: Valid aggregation pipeline stage
change.stream.full.document
Type: string

Description:
Determines what values your change stream returns on update operations.
The default setting returns the differences between the original document and the updated document.
The updateLookup setting returns the differences between the original document and updated document as well as a copy of the entire updated document at a point in time after the update.
The whenAvailable setting returns the updated document, if available.
The required setting returns the updated document and raises an error if it is not available.

Tip

For more information on how this change stream option works, see the MongoDB server manual guide on Lookup Full Document for Update Operations.


Default: ""
Accepted Values: "", "updateLookup", "whenAvailable", or "required"
change.stream.full.document.before.change
Type: string

Description:
Configures the document pre-image your change stream returns on update operations. The pre-image is not available for source records published while copying existing data, and the pre-image configuration has no effect on copying.

Tip

To learn how to configure a collection to enable pre-images, see the Server manual entry on pre- and post-images.

The default setting suppresses the document pre-image.
The whenAvailable setting returns the document pre-image if it's available, before it was replaced, updated, or deleted.
The required setting returns the document pre-image and raises an error if it is not available.

Default: ""
Accepted Values: "" or "whenAvailable" or "required"
publish.full.document.only
Type: boolean

Description:
Whether to return only the fullDocument field from the change stream event document produced by any update event. The fullDocument field contains the most current version of the updated document. To learn more about the fullDocument field, see the Server manual page on update events.
When set to true, the connector overrides the change.stream.full.document setting and sets it to updateLookup so that the fullDocument field contains updated documents.
Default: false
Accepted Values: true or false
publish.full.document.only.tombstones.on.delete
Type: boolean

Description:
Whether to return tombstone events when documents are deleted. Tombstone events contain the keys of deleted documents with null values. This setting applies only when publish.full.document.only is true.
Default: false
Accepted Values: true or false
collation
Type: string

Description:
A JSON collation document that specifies language-specific ordering rules that MongoDB applies to the documents returned by the change stream.

Default: ""
Accepted Values: A valid collation JSON document
batch.size
Type: int

Description:
The change stream cursor batch size.

Default: 0
Accepted Values: An integer
poll.await.time.ms
Type: long

Description:
The amount of time in milliseconds to wait before checking the change stream cursor for new results.

Default: 5000
Accepted Values: An integer
poll.max.batch.size
Type: int

Description:
Maximum number of documents to read in a single batch when polling a change stream cursor for new data. You can use this setting to limit the amount of data buffered internally in the connector.

Default: 1000
Accepted Values: An integer

Use the following configuration settings to specify the format of data the MongoDB Kafka source connector publishes to Kafka topics.

To view only the options related to the format of your output, see the Output Format Properties page.

Name
Description
output.format.key
Type: string

Description:
Specifies which data format the source connector outputs the key document.

Default: json
Accepted Values: bson, json, schema
output.format.value
Type: string

Description:
Specifies which data format the source connector outputs the value document.

Default: json
Accepted Values: bson, json, schema

Note

Protobuf Output Format Support

The connector supports Protobuf as an output data format. You can enable this format by specifying the schema value and installing and configuring the Kafka Connect Protobuf Converter.

output.json.formatter
Type: string

Description:
Class name of the JSON formatter the connector should use to output data.

Default:
com.mongodb.kafka.connect.source.json.formatter.DefaultJson
Accepted Values:
One of the following full class names:
com.mongodb.kafka.connect.source.json.formatter.DefaultJson
com.mongodb.kafka.connect.source.json.formatter.ExtendedJson
com.mongodb.kafka.connect.source.json.formatter.SimplifiedJson
Or your custom JSON formatter full class name.
output.schema.key
Type: string

Description:
Specifies an AVRO schema definition for the key document of the SourceRecord.

Tip

See also:

For more information on AVRO schema, see the Data Formats guide.

Default:
{
"type": "record",
"name": "keySchema",
"fields" : [ { "name": "_id", "type": "string" } ]"
}
Accepted Values: A valid AVRO schema
output.schema.value
Type: string

Description:
Specifies an AVRO schema definition for the value document of the SourceRecord.

Tip

See also:

For more information on AVRO schema, see the Data Formats guide.

Default:
{
"name": "ChangeStream",
"type": "record",
"fields": [
{ "name": "_id", "type": "string" },
{ "name": "operationType", "type": ["string", "null"] },
{ "name": "fullDocument", "type": ["string", "null"] },
{ "name": "ns",
"type": [{"name": "ns", "type": "record", "fields": [
{"name": "db", "type": "string"},
{"name": "coll", "type": ["string", "null"] } ]
}, "null" ] },
{ "name": "to",
"type": [{"name": "to", "type": "record", "fields": [
{"name": "db", "type": "string"},
{"name": "coll", "type": ["string", "null"] } ]
}, "null" ] },
{ "name": "documentKey", "type": ["string", "null"] },
{ "name": "updateDescription",
"type": [{"name": "updateDescription", "type": "record", "fields": [
{"name": "updatedFields", "type": ["string", "null"]},
{"name": "removedFields",
"type": [{"type": "array", "items": "string"}, "null"]
}] }, "null"] },
{ "name": "clusterTime", "type": ["string", "null"] },
{ "name": "txnNumber", "type": ["long", "null"]},
{ "name": "lsid", "type": [{"name": "lsid", "type": "record",
"fields": [ {"name": "id", "type": "string"},
{"name": "uid", "type": "string"}] }, "null"] }
]
}
Accepted Values: A valid JSON schema
output.schema.infer.value
Type: boolean

Description:
Whether the connector should infer the schema for the value document of the SourceRecord. Since the connector processes each document in isolation, the connector may generate many schemas.

Important

The connector only reads this setting when you set your output.format.value setting to schema.

Default: false
Accepted Values: true or false

Use the following configuration settings to configure startup of the MongoDB Kafka source connector to convert MongoDB collections into Change Stream events.

To view only the options related to startup, see the Startup Properties page.

Name
Description
startup.mode
Type: string

Description:
Specifies how the connector should start up when there is no source offset available. Resuming a change stream requires a resume token, which the connector gets from the source offset. If no source offset is available, the connector may either ignore all or some of the existing source data, or may at first copy all existing source data and then continue with processing new data.

If startup.mode=latest, the connector ignores all existing source data.

If startup.mode=timestamp, the connector actuates startup.mode.timestamp.* properties. If no properties are configured, timestamp is equivalent to latest.

If startup.mode=copy_existing, the connector copies all existing source data to Change Stream events. This setting is equivalent to the deprecated setting copy.existing=true.

Note

Data Copy Can Produce Duplicate Events

If any system changes the data in the database while the source connector converts existing data from it, MongoDB may produce duplicate change stream events to reflect the latest changes. Since the change stream events on which the data copy relies are idempotent, the copied data is eventually consistent.

Default:latest
Accepted Values: latest, timestamp, copy_existing
startup.mode.timestamp.start.at.operation.time
Type: string

Description:
Actuated only if startup.mode=timestamp. Specifies the starting point for the change stream. To learn more about Change Stream parameters, see the Server manual entry.
Default: ""
Accepted Values:
  • An integer number of seconds since the Epoch in decimal format (for example, 30)

  • An instant in the ISO-8601 format with one second precision (for example, 1970-01-01T00:00:30Z)

  • A BSON Timestamp in the canonical extended JSON (v2) format (for example, {"$timestamp": {"t": 30, "i": 0}})

startup.mode.copy.existing.namespace.regex
Type: string

Description:
Regular expression the connector uses to match namespaces from which to copy data. A namespace describes the MongoDB database name and collection separated by a period (for example, databaseName.collectionName).

Example

In the following example, the regular-expression setting matches collections that start with "page" in the stats database.

startup.mode.copy.existing.namespace.regex=stats\.page.*

The "" character in the example above escapes the "." character that follows it in the regular expression. For more information on how to build regular expressions, see the Java API documentation on Patterns.

Default: ""
Accepted Values: A valid regular expression
startup.mode.copy.existing.pipeline
Type: string

Description:
An inline array of pipeline operations the connector runs when copying existing data. You can use this setting to filter the source collection and improve the use of indexes in the copying process.

Example

The following example shows how you can use the $match aggregation operator to instruct the connector to copy only documents that contain a closed field with a value of false.

startup.mode.copy.existing.pipeline=[ { "$match": { "closed": "false" } } ]
Default: ""
Accepted Values: Valid aggregation pipeline stages
startup.mode.copy.existing.max.threads
Type: int

Description:
The maximum number of threads the connector can use to copy data.
Default: number of processors available in the environment
Accepted Values: An integer
startup.mode.copy.existing.queue.size
Type: int

Description:
The size of the queue the connector can use when copying data.
Default: 16000
Accepted Values: An integer
startup.mode.copy.existing.allow.disk.use
Type: boolean

Description:
When set to true, the connector uses temporary disk storage for the copy existing aggregation.
Default: true
Accepted Values: true or false

Use the following configuration settings to specify how the MongoDB Kafka source connector behaves when it encounters errors and to specify settings related to resuming interrupted reads.

To view only the options related to handling errors, see the Error Handling and Resuming from Interruption Properties page.

Name
Description
mongo.errors.tolerance
Type: string

Description:
Whether to continue processing messages when the connector encounters an error.
Set this to "none" if you want the connector to stop processing messages and report the issue if it encounters an error.
Set this to "all" if you want the connector to continue processing messages and ignore any errors it encounters.

Important

This property overrides the errors.tolerance Connect Framework property.


Default: "none"
Accepted Values: "none" or "all"
mongo.errors.log.enable
Type: boolean

Description:
Whether the connector should report errors in the log file.
Set this to true to log all errors the connector encounters.
Set this to false to log errors that are not tolerated by the connector. You can specify which errors the connector should tolerate using the errors.tolerance or mongo.errors.tolerance setting.

Important

This property overrides the errors.log.enable Connect Framework property.


Default: false
Accepted Values: true or false
mongo.errors.deadletterqueue.topic.name
Type: string

Description:
The name of topic to use as the dead letter queue.
If you specify a value, the connector writes invalid messages to the dead letter queue topic as extended JSON strings.
If you leave this setting blank, the connector does not write invalid messages to any topic.

Important

You must set errors.tolerance or mongo.errors.tolerance setting to "all" to enable this property.

Default: ""
Accepted Values: A valid Kafka topic name
offset.partition.name
Type: string

Description:
The custom offset partition name to use. You can use this option to instruct the connector to start a new change stream when an existing offset contains an invalid resume token. If you leave this setting blank, the connector uses the default partition name based on the connection details. To view a strategy for naming offset partitions, see the Reset Stored Offsets guide.
Default: ""
Accepted Values: A string. To learn more about naming a partition, see SourceRecord in the Apache Kafka API documentation.
heartbeat.interval.ms
Type: long

Description:
The number of milliseconds the connector waits between sending heartbeat messages. The connector sends heartbeat messages when source records are not published in the specified interval.
Heartbeat messages contain a postBatchResumeToken data field. The value of this field contains the MongoDB server oplog entry that the connector last read from the change stream.
This mechanism improves resumability of the connector for low volume namespaces. See the Invalid Resume Token page for more information on this feature.
Set this to 0 to disable heartbeat messages.
Default: 0
Accepted Values: An integer
heartbeat.topic.name
Type: string

Description:
The name of the topic on which the connector should publish heartbeat messages. You must provide a positive value in the heartbeat.interval.ms setting to enable this feature.

Default: __mongodb_heartbeats
Accepted Values: A valid Kafka topic name
←  Error Handling and Resuming from Interruption PropertiesUsage Examples →
Share Feedback
© 2023 MongoDB, Inc.

About

  • Careers
  • Investor Relations
  • Legal Notices
  • Privacy Notices
  • Security Information
  • Trust Center
© 2023 MongoDB, Inc.