Docs Menu
Docs Home
/
MongoDB Kafka Connector
/ /

Topic Naming

On this page

  • Overview
  • Topic Prefix
  • Topic Suffix
  • Topic Namespace Map
  • Database and Collection Names
  • Regular Expressions
  • Wildcard

The examples on this page show how to configure your MongoDB Kafka source connector to customize the name of the topic to which it publishes records.

By default, the MongoDB Kafka source connector publishes change event data to a Kafka topic with the same name as the MongoDB namespace from which the change events originated. A namespace is a string that's composed of the database and collection name concatenated with a dot (.) character.

The following sections show different ways that you can customize the Kafka topics to which the connector publishes change event data:

  • Topic Prefix

  • Topic Suffix

  • Topic Namespace Map

You can configure your source connector to prepend a string to the namespace of the change event data, and publish records to that Kafka topic. This setting automatically concatenates your prefix with your namespace with the "." character.

To specify the topic prefix, use the topic.prefix configuration setting as shown in the following example:

topic.prefix=myPrefix
database=test
collection=data

Once set, your connector publishes any changes to the data collection in the test database to the Kafka topic named myPrefix.test.data.

You can configure your source connector to append a string to the namespace of the change event data, and publish records to that Kafka topic. This setting automatically concatenates your namespace with your suffix with the "." character.

To specify the topic suffix, use the topic.suffix configuration setting as shown in the following example:

topic.suffix=mySuffix
database=test
collection=data

Once set, your connector publishes any changes to the data collection in the test database to the Kafka topic named test.data.mySuffix.

You can configure your source connector to map namespace values to Kafka topic names for incoming change event data. Topic namespace maps contain pairs that are made up of a namespace pattern and a destination topic name template.

The following sections describe how the connector interprets namespaces and maps them to Kafka topics. In addition to directly mapping databases and collections to Kafka topics, the connector supports the use of regex and wildcard pairs in topic namespace maps.

The order of the pairs in your namespace map can affect how the connector writes events to your topics. The connector matches namespaces in the following order:

  1. Pairs with database and collection names in the namespace pattern. To learn more about this namespace pattern, see the Database and Collection Names example.

  2. Pairs with only a database name in the namespace pattern. To learn more about this namespace pattern, see the Database and Collection Names example.

  3. Regex pairs in order. To learn more about this namespace pattern, see the Regular Expressions example.

  4. The wildcard pair. To learn more about this namespace pattern, see the Wildcard example.

You can specify the names of specific databases and collections within a topic namespace map to write change events from these sources to Kafka topics.

If the database name or namespace of the change event matches one of the fields in the map, the connector computes the destination topic name based on the topic name template that corresponds to that mapping and publishes the event to this topic.

If the database name or namespace of the change event does not match any mapping, the connector publishes the record using the default topic naming scheme unless otherwise specified by a different topic naming setting.

Important

Because the / character denotes that the namespace is a regex pattern, the connector raises a ConnectConfigException if the namespace includes this character in a non-regex context.

Any mapping that includes both database and collection takes precedence over mappings that only specify the source database name.

Important

The namespace map matching occurs before the connector applies any other topic naming setting. If defined, the connector applies the topic.prefix and the topic.suffix settings to the topic name after the mapping.

The following example shows how to specify the topic.namespace.map setting to define a topic namespace mappings from the carDb database to the automobiles topic name template and the carDb.ev namespace to the electricVehicles topic name template:

topic.namespace.map={"carDb": "automobiles", "carDb.ev": "electricVehicles"}

Since the carDb.ev namespace mapping takes precedence over the carDb mapping, the connector performs the following actions:

  • If the change event came from the database carDb and collection ev, the connector sets the destination to the electricVehicles topic.

  • If the change event came from the database carDb and a collection other than ev, the connector sets the destination to the automobiles.<collection name> topic.

  • If the change document came from any database other than carDb, the connector sets the destination topic to the default namespace naming scheme.

  • If you define the topic.prefix and topic.suffix settings, the connector applies their values to the destination topic name after it performs the namespace mapping.

You can use a regular expression (regex) within a topic namespace map. To use a regular expression, start the namespace pattern with the forward slash (/) character. Create the regular expression following the syntax and behavior specified by the java.util.regex.Pattern class.

The connector computes the topic name by doing variable expansion on the topic name template. The connector supports the following variables:

  • db: The database name from the matching namespace.

  • sep: The value of the topic.separator configuration property. To learn more about this property, see Kafka Topic Properties.

  • coll: The collection name from the matching namespace, or an empty string if there is no collection name.

  • sep_coll: The value of the coll variable prefixed with the value of the sep variable, or an empty string if the value of coll is empty.

  • coll_sep: The value of the coll variable suffixed with the value of the sep variable, or an empty string if the value of coll is empty.

  • sep_coll_sep: The value of the coll variable prefixed and suffixed with the value of the sep variable, or an empty string if the value of coll is empty.

If you use any of the previous variables in your topic name template, you must enclose them in curly brackets ({}) for the connector to expand them. You can't use curly brackets in the topic name template for any other purpose.

Note

Escaping Characters

When you create a namespace pattern, ensure that you properly handle characters that need escaping. For example, to match ., the regex syntax requires that you escape it as \. However, you must escape the backslash \ character as \\ according to the JSON syntax. Then, to match . in your namespace pattern, you must write it as \\..

This example shows how to specify the topic.namespace.map setting so that the connector performs the following mapping:

  • Writes events from databases that match a regular expression to a topic computed from the industrial{sep_coll} topic name template. The regex pattern matches any namespace with the vertical database name.

  • Writes events from the vertical.health namespace to the healthcare topic name. In this case, the topic name template and the computed topic name are the same.

topic.namespace.map={"/vertical(?:\\..*)?": "industrial{sep_coll}", "vertical.health": "healthcare"}

Note

In this example, the topic.separator configuration property is ".", the default value.

Because the vertical.health namespace mapping takes precedence over the regular expression namespace mapping, the connector performs the following actions:

  • If the change event comes from the health collection of the vertical database, the connector sets the destination to the healthcare topic.

  • If the change event comes from any other namespaces with the vertical database name, the connector computes the destination topic based on the industrial{sep_coll} topic name template. The following examples demonstrate this mapping:

    • If the change event comes from the vertical.wasteManagement namespace, the connector writes to the industrial.wasteManagement topic.

    • If the change event comes from the vertical database but no specific collection, the connector writes to the industrial topic.

  • If the change document comes from any database that doesn't match the regex pattern, the connector sets the destination topic to the default namespace naming scheme.

  • If you define the topic.prefix and topic.suffix settings, the connector applies their values to the destination topic name after it performs the namespace mapping.

In addition to specifying database name and namespace in your topic namespace map as shown in Database and Collection Names, you can use a wildcard * to match change events from all databases and namespaces without mappings.

topic.namespace.map={"carDb": "automobiles", "carDb.ev": "electricVehicles", "*": "otherVehicles"}

In the preceding wildcard example, the connector publishes change documents that originated from all databases other than carDb to the otherVehicles topic.

Back

Multiple Sources