Updated on 2024-10-08 GMT+08:00

Creating a Kafka Topic

Scenario

You can use the cluster client to create Kafka topics based on service requirements. Management permission is required for clusters with Kerberos authentication enabled.

Prerequisites

You have installed the Kafka client.

Creating a Kafka Topic Using the Kafka Client

  1. Access the ZooKeeper instance page.

    • For versions earlier than MRS 3.x, click the cluster name to go to the cluster details page, and choose Components > ZooKeeper > Instances.

      If the Components tab is unavailable, complete IAM user synchronization first. (On the Dashboard page, click Synchronize on the right side of IAM User Sync to synchronize IAM users.)

    • For MRS 3.x or later, log in to FusionInsight Manager. For details, see Accessing FusionInsight Manager (MRS 3.x or Later). Choose Cluster > Name of the desired cluster > Services > ZooKeeper > Instance.

  2. View the IP addresses of the ZooKeeper role instance.

    Record any IP address of the ZooKeeper instance.

  3. Prepare the client based on service requirements. Log in to the node where the client is installed.

    Log in to the node where the client is installed. For details, see Using an MRS Client. .)

  4. Run the following command to switch to the client installation directory, for example, /opt/client/Kafka/kafka/bin.

    cd /opt/client/Kafka/kafka/bin

  5. Run the following command to configure environment variables:

    source /opt/client/bigdata_env

  6. Run the following command to perform user authentication (skip this step in normal mode):

    kinit Component service user

  7. To perform various operations on Kafka topics for versions prior to MRS 3.x, execute the appropriate commands.

    • Creating a topic

      sh kafka-topics.sh --create --topic Topic name --partitions Number of partitions occupied by the topic --replication-factor Number of replicas of the topic --zookeeper IP address of the node where the ZooKeeper instance resides:clientPort/kafka

    • Deleting a topic

      sh kafka-topics.sh --delete --topic Topic name --zookeeper IP address of the node where the ZooKeeper instance resides:clientPort/kafka

    • The number of topic partitions or topic backup replicas cannot exceed the number of Kafka instances.
    • By default, the value of clientPort of ZooKeeper is 2181.
    • There are three ZooKeeper instances. Use the IP address of any one.
    • For details about managing messages in Kafka topics, see Managing Messages in Kafka Topics.

  8. MRS 3.x and later versions: Use kafka-topics.sh to create Kafka topics.

    • Creating a topic:

      By default, partitions of a topic are distributed based on the number of partitions on the node and disk. To distribute partitions based on the disk capacity, set log.partition.strategy to capacity for the Kafka service.

      When a topic is created in Kafka, partitions and copies can be generated based on the combination of rack awareness and cross-AZ feature. The --zookeeper and --bootstrap-server modes are supported.

      • Disable the rack policy and cross-AZ feature (default policy).

        Copies of topics created based on this policy are randomly allocated to any node in the cluster.

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic--replication-factor number of replicas of the topic--zookeeper IP address of any ZooKeeper node:clientPort/kafka

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic--replication-factor number of replicas of the topic --bootstrap-server IP address of the Kafkacluster:21007 --command-config ../config/client.properties

        If you use --bootstrap-server to create a topic, set rack.aware.enable and az.aware.enable to false.

      • Enable the rack policy and disable the cross-AZ feature.

        The leader of each partition of the topic created based on this policy is randomly allocated on the cluster node. However, different replicas of the same partition are allocated to different racks. Therefore, when this policy is used, ensure that the number of nodes in each rack is the same, otherwise, the load of nodes in the rack with fewer nodes is much higher than the average load of the cluster.

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic--replication-factor number of replicas of the topic --zookeeper IP address of any ZooKeeper node:clientPort/kafka --enable-rack-aware

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic--replication-factor number of replicas of the topic --bootstrap-server IP address of the Kafkacluster:21007 --command-config ../config/client.properties

        If you use --bootstrap-server to create a topic, set rack.aware.enable to true and az.aware.enable to false.

      • Disable the rack policy and enable the cross-AZ feature.

        The leader of each partition of the topic created based on this policy is randomly allocated on the cluster node. However, different replicas of the same partition are allocated to different AZs. Therefore, when this policy is used, ensure that the number of nodes in each AZ is the same, otherwise, the load of nodes in the AZ with fewer nodes is much higher than the average load of the cluster.

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic--replication-factor number of replicas of the topic --zookeeper IP address of any ZooKeeper node:clientPort/kafka --enable-az-aware

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic--replication-factor number of replicas of the topic --bootstrap-server IP address of the Kafkacluster:21007 --command-config ../config/client.properties

        If you use --bootstrap-server to create a topic, set rack.aware.enable to false and az.aware.enable to true.

      • Enable the rack policy and cross-AZ feature.

        The leader of each partition of the topic created based on this policy is randomly allocated on the cluster node. However, different replicas of the same partition are allocated to different racks in different AZs. This policy ensures that the number of nodes on each rack in each AZ is the same, otherwise, the load in the cluster is unbalanced.

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic --replication-factor number of replicas of the topic --zookeeper IP address of any ZooKeeper node:clientPort/kafka --enable-rack-aware --enable-az-aware

        ./kafka-topics.sh --create --topic topic name --partitions number of partitions occupied by the topic--replication-factor number of replicas of the topic --bootstrap-server IP address of the Kafkacluster:21007 --command-config ../config/client.properties

        If you use --bootstrap-server to create a topic, set rack.aware.enable and az.aware.enable to true.

      • Kafka supports topic creation in either of the following modes:
        • In --zookeeper mode, the client generates a copy allocation scheme. The community supports this mode from the beginning. To reduce the dependency on the ZooKeeper component, the community will delete the support for this mode in later versions. When creating a topic in this mode, you can select a copy allocation policy by combining the --enable-rack-aware and --enable-az-aware options. Note: The --enable-az-aware option can be used only when the cross-AZ feature is enabled on the server, that is, az.aware.enable is set to true. Otherwise, the execution fails.
        • In --bootstrap-server mode, the server generates a copy allocation solution. In later versions, the community supports only this mode for topic management. When a topic is created in this mode, the --enable-rack-aware and --enable-az-aware options cannot be used to control the copy allocation policy. The rack.aware.enable and az.aware.enable parameters can be used together to control the copy allocation policy. Note that the az.aware.enable parameter cannot be modified; if the cross-AZ feature is enabled during cluster creation, this parameter is automatically set to true; the rack.aware.enable parameter can be customized.
    • Viewing the topic:
      • ./kafka-topics.sh --describe --zookeeper service IP address of any ZooKeeper node:clientPort/kafka --topic topic name
      • ./kafka-topics.sh --describe --bootstrap-server IP address of the Kafkacluster:21007 --command-config ../config/client.properties --topic topic name