Stream a Changefeed to Amazon MSK Serverless

can stream change data to Amazon MSK Serverless clusters (Amazon Managed Streaming for Apache Kafka), which is an Amazon MSK cluster type that automatically scales your capacity.

MSK Serverless requires IAM authentication for the changefeed to connect to the cluster.

In this tutorial, you’ll set up an MSK Serverless cluster and connect a changefeed with IAM authentication. You’ll create the MSK Serverless cluster with an IAM policy and role. CockroachDB and a Kafka client will assume the IAM role in order to connect to the MSK Serverless cluster. Then you’ll set up the Kafka client to consume the changefeed messages and start the changefeed on the CockroachDB cluster.

Before you begin

You’ll need:

• An AWS account.
• A CockroachDB self-hosted cluster hosted on AWS. You can set up a cluster using . You must create instances in the same VPC that the MSK Serverless cluster will use in order for the changefeed to authenticate successfully.
• A Kafka client to consume the changefeed messages. You must ensure that your client machine is in the same VPC as the MSK Serverless cluster. This tutorial uses a client set up following the AWS MSK Serverless guide.
• The CHANGEFEED privilege in order to create and manage changefeed jobs. Refer to for more details.

You can stream a changefeed to a public IP MSK endpoint from any CockroachDB cluster. If you would like to connect a changefeed running on a CockroachDB Advanced cluster to an Amazon MSK Serverless cluster over AWS PrivateLink, contact your Cockroach Labs account team.

Step 1. Create an MSK Serverless cluster

1. In the AWS Management Console, go to the Amazon MSK console and click Create cluster.
2. If you select Quick create, AWS will provision the cluster in the default VPC for the account. To modify the VPC or network settings, select Custom create. (The VPC selection is important because the MSK Serverless cluster must be in the same VPC as the CockroachDB instance and Kafka client machine.)
3. Ensure you name the cluster and select Serverless as the cluster type. Click Create cluster.
4. Once the cluster is running, click View client information in the Cluster summary box. Copy the endpoint address, which will be similar to boot-vab1abab.c1.kafka-serverless.us-east-1.amazonaws.com:9098. Click Done to return to the cluster’s overview page.

Step 2. Create an IAM policy and role to access the MSK Serverless cluster

MSK Serverless clusters only support IAM authentication. In this step, you’ll create an IAM policy that contains the permissions to interact with the MSK Serverless cluster. Then, you’ll create an IAM role, which you’ll associate with the IAM policy. In a later step, both the CockroachDB cluster and Kafka client machine will use this role to work with the MSK Serverless cluster.

1. In the AWS Management Console, go to the IAM console, select Policies from the navigation, and then Create Policy.

2. Using the JSON tab option, update the policy with the following JSON. These permissions will allow you to connect to the cluster, manage topics, and consume messages. You may want to adjust the permissions to suit your permission model. For more details on the available permissions, refer to the AWS documentation on IAM Access Control for MSK.

   Replace the instances of arn:aws:kafka:{region}:{account ID}:cluster/{msk-cluster-name} with the MSK ARN from your cluster’s summary page and add /* to the end, like the following:

   {
       "Version": "2012-10-17",
       "Statement": [
           {
               "Effect": "Allow",
               "Action": [
                   "kafka-cluster:Connect",
                   "kafka-cluster:AlterCluster",
                   "kafka-cluster:DescribeCluster"
               ],
               "Resource": [
                   "arn:aws:kafka:{region}:{account ID}:cluster/{msk-cluster-name}/*"
               ]
           },
           {
               "Effect": "Allow",
               "Action": [
                   "kafka-cluster:*Topic",
                   "kafka-cluster:WriteData",
                   "kafka-cluster:ReadData"
               ],
               "Resource": [
                   "arn:aws:kafka:{region}:{account ID}:cluster/{msk-cluster-name}/*"
               ]
           },
           {
               "Effect": "Allow",
               "Action": [
                   "kafka-cluster:AlterGroup",
                   "kafka-cluster:DescribeGroup"
               ],
               "Resource": [
                   "arn:aws:kafka:{region}:{account ID}:cluster/{msk-cluster-name}/*"
               ]
           }
       ]
   }
   

3. Once you have added your policy, add a policy name (for example, msk-policy), click Next, and Create policy.

4. Return to the IAM console, select Roles from the navigation, and then Create role.

5. Select AWS service for the Trusted entity type. For Use case, select EC2 from the dropdown. Click Next.

6. On the Add permissions page, search for the IAM policy (msk-policy) you just created. Click Next.

7. Name the role (for example, msk-role) and click Create role.

Step 3. Set up the CockroachDB cluster role

In this step, you’ll create a role, which contains the sts:AssumeRole permission, for the EC2 instance that is running your CockroachDB cluster. The sts:AssumeRole permission will allow the EC2 instance to obtain temporary security credentials to access the MSK Serverless cluster according to the msk-policy permissions. To achieve this, you’ll add the EC2 role to the trust relationship of the msk-role you created in the previous step.

1. Navigate to the IAM console, select Roles from the navigation, and then Create role.

2. Select AWS service for the Trusted entity type. For Use case, select EC2 from the dropdown. Click Next.

3. On the Add permissions page, click Next.

4. Name the role (for example, ec2-role) and click Create role.

5. Once the role has finished creating, copy the ARN in the Summary section. Click on the Trust relationships tab. You’ll find a Trusted entities policy:

   {
       "Version": "2012-10-17",
       "Statement": [
           {
               "Effect": "Allow",
               "Principal": {
                   "Service": "ec2.amazonaws.com"
               },
               "Action": "sts:AssumeRole"
           }
       ]
   }
   

6. Navigate to the IAM console and search for the role (msk-role) you created in Step 2 that contains the MSK policy. Select the role, which will take you to its summary page.

7. Click on the Trust relationships tab, and click Edit trust policy. Add the ARN of the EC2 IAM role (ec2-role) to the JSON policy:

   {
       "Version": "2012-10-17",
       "Statement": [
           {
               "Effect": "Allow",
               "Principal": {
                   "Service": "ec2.amazonaws.com",
                   "AWS": "arn:aws:iam::{account ID}:role/{ec2-role}"
               },
               "Action": "sts:AssumeRole"
           }
       ]
   }
   

   Once you’ve updated the policy, click Update policy.

Step 4. Connect the client to the MSK Serverless cluster

In this step, you’ll prepare the client to connect to the MSK Serverless cluster and create a Kafka topic.

1. Ensure that your client can connect to the MSK Serverless cluster. This tutorial uses an EC2 instance running Kafka as the client. Navigate to the summary page for the client EC2 instance. Click on the Actions dropdown. Click Security, and then select Modify IAM role.

2. On the Modify IAM role page, select the role you created for the MSK Serverless cluster (msk-role) that contains the policy created in Step 2. Click Update IAM role.

3. Open a terminal and connect to your Kafka client. Check that your client.properties file contains the correct SASL and security configuration, like the following:

   security.protocol=SASL_SSL
   sasl.mechanism=AWS_MSK_IAM
   sasl.jaas.config=software.amazon.msk.auth.iam.IAMLoginModule required;
   sasl.client.callback.handler.class=software.amazon.msk.auth.iam.IAMClientCallbackHandler
   

   If you need further detail on setting up the Kafka client, refer to the AWS setup guide.

4. Move to the directory of your Kafka installation:

   cd kafka_2.12-2.8.1/bin
   

5. It is necessary to create topics manually for MSK Serverless clusters. To create a topic, run the following:

   ~/kafka_2.12-2.8.1/bin/kafka-topics.sh --bootstrap-server {msk serverless endpoint} --command-config client.properties --create --topic {users} --partitions {1}
   

   Replace:

   • with your endpoint copied in Step 1.4.
   • with your topic name. This tutorial will use the CockroachDB movr workload and will run a changefeed on the movr.users table.
   • with the number of partitions you require.

   You will receive confirmation output:

   Created topic users.
   

Step 5. Start the changefeed

In this step, you’ll prepare your CockroachDB cluster to start the changefeed.

1. (Optional) On the EC2 instance running CockroachDB, run the application workload to set up some data for your changefeed.

   Create the schema for the workload:

   cockroach workload init movr
   

   Then run the workload:

   cockroach workload run movr --duration=1m
   

2. Start a SQL session. For details on the available flags, refer to the page.

   cockroach sql --insecure
   

To set your Enterprise license, refer to the page.

1. Enable the kv.rangefeed.enabled :

   SET CLUSTER SETTING kv.rangefeed.enabled = true;
   

2. To connect the changefeed to the MSK Serverless cluster, the URI must contain the following parameters:

   • The MSK Serverless cluster endpoint prefixed with the kafka:// scheme, for example: kafka://boot-vab1abab.c1.kafka-serverless.us-east-1.amazonaws.com:9098.
   • tls_enabled set to true.
   • sasl_enabled set to true.
   • sasl_mechanism set to AWS_MSK_IAM.
   • sasl_aws_region set to the region of the MSK Serverless cluster.
   • sasl_aws_iam_role_arn set to the ARN for the IAM role (msk-role) that has the permissions outlined in Step 2.2.
   • sasl_aws_iam_session_name set to a string that you specify to identify the session in AWS.
   'kafka://boot-vab1abab.c1.kafka-serverless.us-east-1.amazonaws.com:9098/?tls_enabled=true&sasl_enabled=true&sasl_mechanism=AWS_MSK_IAM&sasl_aws_region=us-east-1&sasl_aws_iam_role_arn=arn:aws:iam::{account ID}:role/{msk-role}&sasl_aws_iam_session_name={user-specified session name}'
   

   You can either specify the Kafka URI in the CREATE CHANGEFEED statement directly. Or, create an for the MSK Serverless URI.

   External connections define a name for an external connection while passing the provider URI and query parameters:

   CREATE EXTERNAL CONNECTION msk_serverless AS 'kafka://boot-vab1abab.c1.kafka-serverless.us-east-1.amazonaws.com:9098/?tls_enabled=true&sasl_enabled=true&sasl_mechanism=AWS_MSK_IAM&sasl_aws_region=us-east-1&sasl_aws_iam_role_arn=arn:aws:iam::{account ID}:role/{msk-role}&sasl_aws_iam_session_name={user-specified session name}';
   

3. Use the statement to start the changefeed using either the external connection (external://) or full kafka:// URI:

   CREATE CHANGEFEED FOR TABLE movr.users INTO `external://msk_serverless` WITH resolved;
   
           job_id
   -----------------------
   1002677216020987905
   

   To view a changefeed job, use .

Step 6. Consume the changefeed messages on the client

1. Return to the terminal that is running the Kafka client. Move to the Kafka installation directory:

   cd kafka_2.12-2.8.1/bin
   

2. Run the following command to start a consumer. Set --topic to the topic you created in Step 4.5:

   ~/kafka_2.12-2.8.1/bin/kafka-console-consumer.sh --bootstrap-server {msk serverless endpoint} --consumer.config client.properties --topic users --from-beginning
   
   {"after": {"address": "49665 Green Highway", "city": "amsterdam", "credit_card": "0762957951", "id": "10acc68c-5cea-4d32-95db-3254b8a1170e", "name": "Jimmy Gutierrez"}}
   {"after": {"address": "1843 Brandon Common Apt. 68", "city": "amsterdam", "credit_card": "3414699744", "id": "53d95b9a-abf3-4af2-adc8-92d4ee026327", "name": "James Hunt"}}
   {"after": {"address": "87351 David Ferry Suite 24", "city": "amsterdam", "credit_card": "7689751883", "id": "58f66df9-e2ef-48bf-bdbe-436e8caa0fae", "name": "Grant Murray"}}
   {"after": {"address": "35991 Tran Flats", "city": "amsterdam", "credit_card": "6759782818", "id": "6e8d430d-9a3b-4519-b7ab-987d21043f6a", "name": "Mr. Alan Powers"}}
   {"after": {"address": "65320 Emily Ports", "city": "amsterdam", "credit_card": "7424361516", "id": "74e8c91b-9534-4e40-9d19-f23e14d24114", "name": "Michele Grant"}}
   {"after": {"address": "85363 Gary Square Apt. 39", "city": "amsterdam", "credit_card": "0267354734", "id": "99d2c816-2216-40f3-b60c-f19bc3e9f455", "name": "Mrs. Wendy Miller"}}
   {"after": {"address": "68605 Shane Shores Suite 22", "city": "amsterdam", "credit_card": "5913104602", "id": "ae147ae1-47ae-4800-8000-000000000022", "name": "Crystal Sullivan"}}
   {"after": {"address": "41110 Derrick Walk Apt. 42", "city": "amsterdam", "credit_card": "2866469885", "id": "b3333333-3333-4000-8000-000000000023", "name": "Michael Lewis"}}
   {"after": {"address": "47781 Robinson Villages Apt. 41", "city": "amsterdam", "credit_card": "6596967781", "id": "b40e7c51-7e68-43f1-a4df-92fa8a05c961", "name": "Abigail Sellers"}}
   ...
   {"resolved":"1725982589714395510.0000000000"}
   ...
   

See also

For more resources, refer to the following:

• page for details on parameters that sinks support.
• Configuration for serverless clusters in the AWS documentation for details on topic-level configuration options.
• Monitoring serverless clusters in the AWS documentation for details on monitoring the MSK Serverless cluster.
• for details on monitoring the changefeed job.