1. Home
2. Connect External Systems to Confluent Cloud

Azure Cognitive Search Sink Connector for Confluent Cloud

The Azure Cognitive Search Sink connector for Confluent Cloud can move data from Apache Kafka® to Azure Cognitive Search. The connector writes each event from a Kafka topic (as a document) to an index in Azure Cognitive Search. The connector uses the Azure Cognitive Search REST API to send records as documents.

Features

The Azure Cognitive Search Sink connector supports the following features:

• At least once delivery: This connector guarantees that records from the Kafka topic are delivered at least once.

• Supports multiple tasks: The connector supports running one or more tasks. More tasks may improve performance.

• Ordered writes: The connector writes records in exactly the same order that it receives them. And for uniqueness, the Kafka coordinates (topic, partition, and offset) can be used as the document key. Otherwise, the connector uses the record key as the document key.

• Automatically creates topics: The following three topics are automatically created when the connector starts:

  • Success topic
  • Error topic
  • Dead letter queue (DLQ) topic

  The suffix for each topic name is the connector’s logical ID. In the example below, there are the three connector topics and one pre-existing Kafka topic named pageviews.

  

  Connector Topics

  If the records sent to the topic are not in the correct format, or if important fields are missing in the record, the errors are recorded in the error topic, and the connector continues to run.

• Automatic retries: The connector will retry all requests (that can be retried) when the Azure Cognitive Search service is unavailable. The maximum amount of time that the connector spends retrying can be specified by the max.retry.ms configuration property.

• Supported data formats: The connector supports Avro, JSON Schema (JSON-SR), and Protobuf input formats. Schema Registry must be enabled to use these Schema Registry-based formats.

See Configuration Properties for configuration property values and descriptions. See Cloud connector limitations for more information.

Azure service principal

You need an Azure RBAC service principal to run the connector. When you create the service principal, the output from the Azure CLI command provides the necessary authentication and authorization details that you add to the connector configuration.

Note

If you want to assign roles to an existing service principal using the Azure portal instead of the CLI, see Assign Azure roles using the Azure portal.

Complete the following steps to create the service principal using the Azure CLI.

1. Log in to the Azure CLI.

   az login
   

2. Enter the following command to create the service principal:

   az ad sp create-for-rbac --name <Name of service principal> --scopes \
   /subscriptions/<SubscriptionID>/resourceGroups/<Resource_Group>
   

   For example:

   az ad sp create-for-rbac --name azure_search --scopes /subscriptions/
   d92eeba4-...omitted...-37c2bd9259d0/resourceGroups/connect-azure
   
   Creating 'Contributor' role assignment under scope '/subscriptions/
   d92eeba4-...omitted...-37c2bd9259d0/resourceGroups/connect-azure'
   
   The output includes credentials that you must protect. Be sure that you
   do not include these credentials in your code or check the credentials
   into your source control.
   
   {
      "appId": "8ec186f9-...omitted...-e575b928b00a",
      "displayName": "azure_search",
      "name": "8ec186f9-...omitted...-e575b928b00a",
      "password": "jdGzGTwCKQ...omitted...QwE3hx",
      "tenant": "0893715b-...omitted...-2789e1ead045"
   }
   

   Save the following details to use in the connector configuration:

   • Use the "appId" output for the connector UI field named Azure Client ID (CLI property azure.search.client.id).

   • Use the "password" output for the connector UI field named Azure Client Secret (CLI property azure.search.client.secret).

   • Use the "tenant" output for the connector UI field named Azure Tenant ID (CLI property azure.search.tenant.id).

     Tip

     You can create a more granular service principal if needed. For example, the following command creates the contributor role assignment specifically to access Azure search services.

     az ad sp create-for-rbac --name <Name of service principal> --scopes
      /subscriptions/<SubscriptionID>/resourceGroups/<Resource Group>
      /providers/Microsoft.Search/searchServices/<Search Service Name>
      --role Reader
     

Quick Start

Use this quick start to get up and running with the Confluent Cloud Azure Cognitive Search Sink connector. The quick start provides the basics of selecting the connector and configuring it to stream events.

Prerequisites

• Authorized access to a Confluent Cloud cluster on Microsoft Azure (Azure).
• An Azure service principal, an Azure Cognitive Search API key, and subscription details for the connector configuration.
• The Confluent CLI installed and configured for the cluster. See Install the Confluent CLI.
• Schema Registry must be enabled to use a Schema Registry-based format (for example, Avro, JSON_SR (JSON Schema), or Protobuf). See Environment Limitations for additional information.
• At least one index must exist in Azure Cognitive Search.
• All record schema fields must be present as Azure search service index fields.
• At least one source Kafka topic must exist in your Confluent Cloud cluster before creating the sink connector.

Using the Confluent Cloud Console

Step 1: Launch your Confluent Cloud cluster.

See the Quick Start for Apache Kafka using Confluent Cloud for installation instructions.

Step 2: Add a connector.

In the left navigation menu, click Data integration, and then click Connectors. If you already have connectors in your cluster, click + Add connector.

Step 3: Select your connector.

Click the Azure Cognitive Search Sink connector icon.

Step 4: Set up the connection.

Note

• Make sure you have all your prerequisites completed.
• An asterisk ( * ) designates a required entry.

1. Select one or more topics.

2. Enter a connector Name.

3. Select the Input Kafka record value format (data coming from the Kafka topic): AVRO, JSON_SR (JSON Schema), or PROTOBUF. A valid schema must be available in Schema Registry to use a schema-based message format (for example, Avro, JSON_SR (JSON Schema) or Protobuf).

4. Select the way you want to provide Kafka Cluster credentials. You can either select a service account resource ID or you can enter an API key and secret (or generate these in the Cloud Console).

5. Enter the Azure Search service connection details. See Azure service principal and Azure Cognitive Search API key for property details.

6. Enter the Search Service Write Details details.

   • Enter the Index Name Pattern. This is the name of the search index to write records to (as documents). This index must exist in the search service.
   
   

   You can enter any of the remaining optional properties.

   • Write Method: The method that the connector uses to write Kafka records to the search index. The option upload functions like an upsert. That is, a document is inserted if it does not exist, or updated (or replaced) if the document does exist. If you select the option mergeOrUpload, the connector updates specified fields in an existing document. If the document does not exist, then mergeOrUpload functions like upload. The default setting is upload.
   • Delete Enabled: Set whether a document is deleted for a null record. The default option is false.
   • Key Mode: Select an option for the key ID. If you select the option KEY, the connector uses the Kafka record key for the document key. If you select the option COORDINATES, the Kafka coordinates (topic, partition, and offset) are concatenated and used for the document key. The default is KEY.
   • Max Batch Size: Enter the maximum number of records that the connector can send in a batch. The default setting is 1 for no batching.
   • Max Retry Time (ms): The maximum amount of time in milliseconds (ms) that the connector will retry a request before stopping. The default setting is 300000 ms (5 minutes).

7. Enter the number of tasks to use with the connector. More tasks may improve performance.

8. Transforms and Predicates: See the Single Message Transforms (SMT) documentation for details. See Unsupported transformations for a list of SMTs that are not supported with this connector.

See Configuration Properties for configuration property values and descriptions.

Step 5: Launch the connector.

Verify the connection details and click Launch.

Step 6: Check the connector status.

The status for the connector should go from Provisioning to Running.

Step 7: Check for documents.

Verify that documents are populating the search index.

For more information and examples to use with the Confluent Cloud API for Connect, see the Confluent Cloud API for Connect section.

Tip

When you launch a connector, a Dead Letter Queue topic is automatically created. See Dead Letter Queue for details.

Using the Confluent CLI

Complete the following steps to set up and run the connector using the Confluent CLI.

Note

• Make sure you have all your prerequisites completed.
• The example commands use Confluent CLI version 2. For more information see, Confluent CLI v2.

Step 1: List the available connectors.

Enter the following command to list available connectors:

confluent connect plugin list

Step 2: Show the required connector configuration properties.

Enter the following command to show the required connector properties:

confluent connect plugin describe <connector-catalog-name>

For example:

confluent connect plugin describe AzureCognitiveSearchSink

Example output:

Following are the required configs:
connector.class: AzureCognitiveSearchSink
input.data.format
name
kafka.api.key
kafka.api.secret
azure.search.service.name
azure.search.api.key
azure.search.client.id
azure.search.client.secret
azure.search.tenant.id
azure.search.subscription.id
azure.search.resourcegroup.name
index.name
tasks.max
topics

Step 3: Create the connector configuration file.

Create a JSON file that contains the connector configuration properties. The following example shows the required connector properties.

{
  "connector.class": "AzureCognitiveSearchSink",
  "input.data.format": "AVRO",
  "name": "AzureCognitiveSearchSink_0",
  "kafka.api.key": "****************",
  "kafka.api.secret": "************************************************",
  "azure.search.service.name": "<service_name>",
  "azure.search.api.key": "<api_key>",
  "azure.search.client.id": "<client_id>",
  "azure.search.client.secret": "<client_secret>",
  "azure.search.tenant.id": "<tenant_id>",
  "azure.search.subscription.id": "<subscription_id>",
  "azure.search.resourcegroup.name": "<resource_group>",
  "index.name": "<index_name>",
  "tasks.max": "1",
  "topics": "<topic_name>"
}

Note the following property definitions:

• "connector.class": Identifies the connector plugin name.
• "input.data.format": Sets the input Kafka record value format (data coming from the Kafka topic). Valid entries are AVRO, JSON_SR, and PROTOBUF. You must have Confluent Cloud Schema Registry configured if using a schema-based message format (for example, Avro, JSON_SR (JSON Schema), or Protobuf).
• "name": Sets a name for your new connector.
• "kafka.api.key" and ""kafka.api.secret": These credentials are either the cluster API key and secret or the service account API key and secret.
• azure.search.<...> Required Azure and Azure search connection details. See Azure service principal and Azure Cognitive Search API key for property details.
• "index.name": The name of the search index to write records to (as documents).
• "tasks.max": Enter the maximum number of tasks for the connector to use. More tasks may improve performance.
• "topics": Enter the topic name or a comma-separated list of topic names.

Single Message Transforms: See the Single Message Transforms (SMT) documentation for details about adding SMTs using the CLI. See Unsupported transformations for a list of SMTs that are not supported with this connector.

See Configuration Properties for additional configuration property values and descriptions.

Step 4: Load the properties file and create the connector.

Enter the following command to load the configuration and start the connector:

confluent connect create --config <file-name>.json

For example:

confluent connect create --config azure-search-sink-config.json

Example output:

Created connector AzureCognitiveSearchSink_0 lcc-do6vzd

Step 5: Check the connector status.

Enter the following command to check the connector status:

confluent connect list

Example output:

ID           |             Name             | Status  | Type | Trace
+------------+------------------------------+---------+------+-------+
lcc-do6vzd   | AzureCognitiveSearchSink_0   | RUNNING | sink |       |

Step 6: Check for documents.

Verify that the Azure search index is being populated.

For more information and examples to use with the Confluent Cloud API for Connect, see the Confluent Cloud API for Connect section.

Tip

When you launch a connector, a Dead Letter Queue topic is automatically created. See Dead Letter Queue for details.

Configuration Properties

The following connector configuration properties can be used with the Azure Cognitive Search Sink connector for Confluent Cloud.

azure.search.service.name

The name of the Azure Cognitive Search service.

• Type: string
• Importance: high

azure.search.api.key

The API key for the Azure Cognitive Search service.

• Type: password
• Importance: high

azure.search.client.id

The service principal client ID for your subscription.

• Type: password
• Importance: high

azure.search.client.secret

The service principal client ID secret for your subscription.

• Type: password
• Importance: high

azure.search.tenant.id

The service principal tenant ID for your subscription.

• Type: password
• Importance: high

azure.search.subscription.id

The subscription ID for your Azure account.

• Type: password
• Importance: high

azure.search.resourcegroup.name

The resource group where the Azure search service exists.

• Type: string
• Importance: high

index.name

The name of the search index to write records to (as documents).

• Type: string
• Importance: high

write.method

The method used to write Kafka records to the search index.The option upload (the default) functions like an upsert. That is, a document is inserted if it does not exist and updated (or replaced) if it does exist. The option mergeOrUpload updates an existing document with the specified fields. If the document doesn’t exist, this option behaves like upload.

• Type: string
• Default value: upload
• Importance: high

delete.enabled

Sets whether the connector deletes documents if the record value is null.

• Type: string
• Default value: false
• Importance: high

key.mode

Determines what is used for the document key ID. Using KEY designates the Kafka record key for the document key ID. Using COORDINATES designates that the Kafka coordinates (topic, partition, and offset) are concatenated and used for the document key ID. The default is KEY.

• Type: string
• Default value: KEY
• Importance: medium

max.batch.size

The maximum number of Kafka records sent per request. The default value 1 disables record batching.

• Type: int
• Default value: 1
• Importance: high

max.retry.ms

The maximum amount of time in milliseconds (ms) that the connector will retry a request before stopping. The default setting is 300000 ms (5 minutes).

• Type: int
• Default value: 300000
• Importance: high

Next Steps

See also

For an example that shows fully-managed Confluent Cloud connectors in action with Confluent Cloud ksqlDB, see the Cloud ETL Demo. This example also shows how to use Confluent CLI to manage your resources in Confluent Cloud.