Docker Configuration ¶

You can install the Confluent Platform using Docker images. This section provides an overview of Confluent’s Docker images for the Confluent Platform.

Contents

Docker Configuration

Confluent Docker Images ¶

The Confluent Platform Docker images support passing configuration variables dynamically using environment variables. More specifically, we use the Docker -e or --env flags for setting various settings in the respective images when starting up the images.

The images are available for Confluent Platform 3.0.1 and greater. Images are available on DockerHub for each component of the Confluent Platform. Alternatively, the source files for the images are available on Github if you would prefer to extend and/or rebuild the images and upload them to your own DockerHub repository.

The table below lists the available images and the Confluent software packages they contain. You’ll note that some images are identified as `cp-enterprise-${component_name}`. These images include proprietary components that must be licensed from Confluent when deployed.

Component	Image Name	Type	Packages Included
Base Image	cp-base	Open Source	zulu-openjdk-8
Kafka	cp-kafka	Open Source	confluent-kafka-*
Kafka	cp-enterprise-kafka	Enterprise	confluent-kafka-* confluent-rebalancer confluent-support-metrics
Control Center	cp-enterprise-control-center	Enterprise	confluent-control-center
Replicator	cp-enterprise-replicator	Enterprise	confluent-kafka-replicator confluent-schema-registry confluent-control-center
Kafka Connect	cp-kafka-connect	Enterprise	confluent-kafka-connect-jdbc confluent-kafka-connect-hdfs confluent-schema-registry confluent-control-center confluent-kafka-connect-elasticsearch confluent-kafka-connect-s3
Schema Registry	cp-schema-registry	Open Source	confluent-schema-registry
REST Proxy	cp-kafka-rest	Open Source	confluent-kafka-rest

Note: The Kafka Connect image is labeled as “Enterprise” simply because it contains the Confluent Control Center package. That package enables the deployed connectors to collect the metrics visualized in Confluent Control Center. No explicit license is required when using the Kafka Connect image on its own.

Configuration Notes ¶

Docker for Mac
We do not recommend using these images with Docker for Mac at this time. The primary reason is that Docker for Mac does not update the local /etc/hosts file with the hostnames of the deployed containers. This makes it difficult to access the containerized cluster with client applications running directly on the Mac. Additionally, the semantics of --net=host are not clear, so deploying containers with host networking on Docker for Mac is not reliable. More details on these issues can be found at:
- Hostname Issue
- Host networking on Docker for Mac: link 1, link 2, link 3
Persistent Data (Mounted Volumes)

When deploying the Kafka and ZooKeeper images, you should always use mounted volumes for the file systems those images use for their persistent data. This ensures that the containers will retain their proper state when stopped and restarted. The other images maintain their state directly in Kafka topics, so mounted volumes are not usually required for those containers.
Bridge Networking vs. Host Networking
Bridge networking is currently only supported on a single host. For multiple hosts, you will need to use overlay networks which are not currently supported. To expose Kafka to clients outside of the bridge network, you need to find the container IP and put it in advertised.listeners. This can be difficult to achieve depending on how you’re using the images. Furthermore, it can add a network hop and may not be as performant as the host network, which shares the network stack. In summary, host networking is the recommended option in the following cases:
Multi-host clusters without using Swarm/Kubernetes host network is the best approach

If you need clients to be able to access Kafka outside the bridge/overlay network
Launch Settings

Docker containers should be launched with Restart=always unless you are using a process manager. This ensures that intermittent failures in the Docker environment do not result in unnecessary failures of the Confluent services.
Adding Connectors to the Kafka Connect Image
There are currently two ways to add new connectors to the Kafka Connect image.
- Build a new Docker image that has the connector installed. You can follow the examples found in Extending Images. You will need to make sure that the connector jars are on the CLASSPATH for the Connect service (the default location of /usr/share/java/kafka-connect-* is the recommended location).
- Add the connector jars via volumes. If you don’t want to create a new Docker image, please see our documentation on Configuring Kafka Connect with External Jars to configure the cp-kafka-connect container with external jars.
Included Java

The Confluent Docker images are tested and shipped with Azul Zulu OpenJDK. Other JDK’s (including Oracle Java) are supported, but you must extend the images yourself to implement that change.
Untested Features
The following features/environments are not currently tested:
Schema Registry SSL

Kafka Connect with Security Enabled

Confluent Control Center with Security Enabled

The images are not currently tested on Docker Swarm.

Configuration Parameters ¶

Some configuration variables are required when starting up the Docker images. We have outlined those variables below for each component along with an example of how to pass them. For a full list of all available configuration options for each Confluent Platform component, you should refer to their respective documentation.

ZooKeeper ¶

The ZooKeeper image uses variables prefixed with ZOOKEEPER_ with the variables expressed exactly as they would appear in the zookeeper.properties file. As an example, to set clientPort, tickTime, and syncLimit run the command below:

docker run -d \
--net=host \
--name=zookeeper \
-e ZOOKEEPER_CLIENT_PORT=32181 \
-e ZOOKEEPER_TICK_TIME=2000 \
-e ZOOKEEPER_SYNC_LIMIT=2 \
confluentinc/cp-zookeeper:4.0.0

Required Settings¶

ZOOKEEPER_CLIENT_PORT

This field is always required. Tells ZooKeeper where to listen for connections by clients such as Kafka.

ZOOKEEPER_SERVER_ID

Only required when running in clustered mode. Sets the server ID in the myid file, which consists of a single line containing only the text of that machine’s id. So myid of server 1 would contain the text “1” and nothing else. The id must be unique within the ensemble and should have a value between 1 and 255.

Confluent Kafka (cp-kafka)¶

The Kafka image uses variables prefixed with KAFKA_ with an underscore (_) separating each word instead of periods. As an example, to set broker.id, advertised.listeners, zookeeper.connect, and offsets.topic.replication.factor, you’d run the following command:

docker run -d \
    --net=host \
    --name=kafka \
    -e KAFKA_ZOOKEEPER_CONNECT=localhost:32181 \
    -e KAFKA_ADVERTISED_LISTENERS=PLAINTEXT://localhost:29092 \
    -e KAFKA_BROKER_ID=2 \
    -e KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR=1 \
    confluentinc/cp-kafka:4.0.0
Note

You’ll notice that we set the KAFKA_ADVERTISED_LISTENERS variable to localhost:29092. This is an important setting, as it will make Kafka accessible from outside the container by advertising it’s location on the Docker host.

Also notice that we set KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR to 1. This is needed when you are running with a single-node cluster. If you have three or more nodes, you do not need to change this from the default.

Required Settings¶

KAFKA_ZOOKEEPER_CONNECT

Tells Kafka how to get in touch with ZooKeeper.

KAFKA_ADVERTISED_LISTENERS

Advertised listeners is required for starting up the Docker image because it is important to think through how other clients are going to connect to kafka. In a Docker environment, you will need to make sure that your clients can connect to Kafka and other services. Advertised listeners is how it gives out a host name that can be reached by the client.

Confluent Enterprise Kafka (cp-enterprise-kafka)¶

The Enterprise Kafka image includes the packages for Confluent Auto Data Balancing and Proactive support in addition to Kafka. The Enterprise Kafka image uses variables prefixed with KAFKA_ for Apache Kafka and with CONFLUENT_ for Confluent components. These variables have an underscore (_) separating each word instead of periods. As an example, to set broker.id, advertised.listeners, zookeeper.connect, offsets.topic.replication.factor, and confluent.support.customer.id you’d run the following command:

docker run -d \
    --net=host \
    --name=kafka \
    -e KAFKA_ZOOKEEPER_CONNECT=localhost:32181 \
    -e KAFKA_ADVERTISED_LISTENERS=PLAINTEXT://localhost:29092 \
    -e KAFKA_BROKER_ID=2 \
    -e KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR=1 \
    -e CONFLUENT_SUPPORT_CUSTOMER_ID=c0 \
    confluentinc/cp-enterprise-kafka:4.0.0
Note

You’ll notice that we set the KAFKA_ADVERTISED_LISTENERS variable to localhost:29092. This is an important setting, as it will make Kafka accessible from outside the container by advertising it’s location on the Docker host.

If you want to enable Proactive support or use Confluent Auto Data Balancing features, please follow the Proactive support and ADB documentation at Confluent documentation.

Also notice that we set KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR to 1. This is needed when you are running with a single-node cluster. If you have three or more nodes, you do not need to change this from the default.

Required Settings¶

KAFKA_ZOOKEEPER_CONNECT

Tells Kafka how to get in touch with ZooKeeper.

KAFKA_ADVERTISED_LISTENERS

Advertised listeners is required for starting up the Docker image because it is important to think through how other clients are going to connect to kafka. In a Docker environment, you will need to make sure that your clients can connect to Kafka and other services. Advertised listeners is how it gives out a host name that can be reached by the client.

Schema Registry ¶

For the Schema Registry image, use variables prefixed with SCHEMA_REGISTRY_ with an underscore (_) separating each word instead of periods. As an example, to set kafkastore.connection.url, host.name, listeners and debug you’d run the following:

docker run -d \
  --net=host \
  --name=schema-registry \
  -e SCHEMA_REGISTRY_KAFKASTORE_CONNECTION_URL=localhost:32181 \
  -e SCHEMA_REGISTRY_HOST_NAME=localhost \
  -e SCHEMA_REGISTRY_LISTENERS=http://localhost:8081 \
  -e SCHEMA_REGISTRY_DEBUG=true \
  confluentinc/cp-schema-registry:4.0.0

Required Settings¶

SCHEMA_REGISTRY_KAFKASTORE_CONNECTION_URL

ZooKeeper URL for the Kafka cluster.

SCHEMA_REGISTRY_HOST_NAME

The host name advertised in ZooKeeper. Make sure to set this if running Schema Registry with multiple nodes. Hostname is required because it defaults to the Java canonical host name for the container, which may not always be resolvable in a Docker environment. Hostname must be resolveable because slave nodes serve registration requests indirectly by simply forwarding them to the current master, and returning the response supplied by the master. For more information, please refer to the Schema Registry documentation on Single Master Architecture.

Kafka REST Proxy ¶

For the Kafka REST Proxy image use variables prefixed with KAFKA_REST_ with an underscore (_) separating each word instead of periods. As an example, to set the listeners, schema.registry.url and zookeeper.connect you’d run the following command:

docker run -d \
  --net=host \
  --name=kafka-rest \
  -e KAFKA_REST_ZOOKEEPER_CONNECT=localhost:32181 \
  -e KAFKA_REST_LISTENERS=http://localhost:8082 \
  -e KAFKA_REST_SCHEMA_REGISTRY_URL=http://localhost:8081 \
  confluentinc/cp-kafka-rest:4.0.0

Required Settings¶

The following settings must be passed to run the REST Proxy Docker image.

KAFKA_REST_HOST_NAME

The host name used to generate absolute URLs in responses. Hostname is required because it defaults to the Java canonical host name for the container, which may not always be resolvable in a Docker environment. For more details, please refer to the Confluent Platform documentation on REST proxy deployment.

KAFKA_REST_ZOOKEEPER_CONNECT

Specifies the ZooKeeper connection string in the form hostname:port where host and port are the host and port of a ZooKeeper server. To allow connecting through other ZooKeeper nodes when that ZooKeeper machine is down you can also specify multiple hosts in the form hostname1:port1,hostname2:port2,hostname3:port3.

The server may also have a ZooKeeper chroot path as part of it’s ZooKeeper connection string which puts its data under some path in the global ZooKeeper namespace. If so the consumer should use the same chroot path in its connection string. For example to give a chroot path of /chroot/path you would give the connection string as hostname1:port1,hostname2:port2,hostname3:port3/chroot/path.

Kafka Connect ¶

The Kafka Connect image uses variables prefixed with CONNECT_ with an underscore (_) separating each word instead of periods. As an example, to set the required properties like bootstrap.servers, the topic names for config, offsets and status as well the key or value converter, run the following command:

docker run -d \
  --name=kafka-connect \
  --net=host \
  -e CONNECT_BOOTSTRAP_SERVERS=localhost:29092 \
  -e CONNECT_REST_PORT=28082 \
  -e CONNECT_GROUP_ID="quickstart" \
  -e CONNECT_CONFIG_STORAGE_TOPIC="quickstart-config" \
  -e CONNECT_OFFSET_STORAGE_TOPIC="quickstart-offsets" \
  -e CONNECT_STATUS_STORAGE_TOPIC="quickstart-status" \
  -e CONNECT_KEY_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_VALUE_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_INTERNAL_KEY_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_INTERNAL_VALUE_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_REST_ADVERTISED_HOST_NAME="localhost" \
  confluentinc/cp-kafka-connect:4.0.0

Required Settings¶

The following settings must be passed to run the Kafka Connect Docker image.

CONNECT_BOOTSTRAP_SERVERS

A unique string that identifies the Connect cluster group this worker belongs to.

CONNECT_GROUP_ID

A unique string that identifies the Connect cluster group this worker belongs to.

CONNECT_CONFIG_STORAGE_TOPIC

The name of the topic in which to store connector and task configuration data. This must be the same for all workers with the same group.id

CONNECT_OFFSET_STORAGE_TOPIC

The name of the topic in which to store offset data for connectors. This must be the same for all workers with the same group.id

CONNECT_STATUS_STORAGE_TOPIC

The name of the topic in which to store state for connectors. This must be the same for all workers with the same group.id

CONNECT_KEY_CONVERTER

Converter class for keys. This controls the format of the data that will be written to Kafka for source connectors or read from Kafka for sink connectors.

CONNECT_VALUE_CONVERTER

Converter class for values. This controls the format of the data that will be written to Kafka for source connectors or read from Kafka for sink connectors.

CONNECT_INTERNAL_KEY_CONVERTER

Converter class for internal keys that implements the Converter interface.

CONNECT_INTERNAL_VALUE_CONVERTER

Converter class for internal values that implements the Converter interface.

CONNECT_REST_ADVERTISED_HOST_NAME

Advertised host name is required for starting up the Docker image because it is important to think through how other clients are going to connect to Connect REST API. In a Docker environment, you will need to make sure that your clients can connect to Connect and other services. Advertised host name is how Connect gives out a host name that can be reached by the client.

Optional Settings¶

All other settings for Connect like security, monitoring interceptors, producer and consumer overrides can be passed to the Docker images as environment variables. The names of these environment variables are derived by replacing . with _, converting the resulting string to uppercase and prefixing it with CONNECT_. For example, if you need to set ssl.key.password, the environment variable name would be CONNECT_SSL_KEY_PASSWORD.

The image will then convert these environment variables to corresponding Connect config variables.

Confluent Control Center ¶

The Confluent Control Center image uses variables prefixed with CONTROL_CENTER_ with an underscore (_) separating each word instead of periods. As an example, the following command runs Control Center, passing in its ZooKeeper, Kafka, and Connect configuration parameters.

docker run -d \
  --net=host \
  --name=control-center \
  --ulimit nofile=16384:16384 \
  -e CONTROL_CENTER_ZOOKEEPER_CONNECT=localhost:32181 \
  -e CONTROL_CENTER_BOOTSTRAP_SERVERS=localhost:29092 \
  -e CONTROL_CENTER_REPLICATION_FACTOR=1 \
  -e CONTROL_CENTER_CONNECT_CLUSTER=http://localhost:28082 \
  -v /mnt/control-center/data:/var/lib/confluent-control-center \
  confluentinc/cp-enterprise-control-center:4.0.0

Docker Options¶

File descriptor limit: Control Center may require many open files so we recommend setting the file descriptor limit to at least 16384
Data persistence: the Control Center image stores its data in the /var/lib/confluent-control-center directory. We recommend that you bind this to a volume on the host machine so that data is persisted across runs.

Required Settings¶

The following settings must be passed to run the Confluent Control Center image.

CONTROL_CENTER_ZOOKEEPER_CONNECT

Specifies the ZooKeeper connection string in the form hostname:port where host and port are the host and port of a ZooKeeper server. To allow connecting through other ZooKeeper nodes when that ZooKeeper machine is down you can also specify multiple hosts in the form hostname1:port1,hostname2:port2,hostname3:port3.

The server may also have a ZooKeeper chroot path as part of it’s ZooKeeper connection string which puts its data under some path in the global ZooKeeper namespace. If so the consumer should use the same chroot path in its connection string. For example to give a chroot path of /chroot/path you would give the connection string as hostname1:port1,hostname2:port2,hostname3:port3/chroot/path.

CONTROL_CENTER_BOOTSTRAP_SERVERS

A list of host/port pairs to use for establishing the initial connection to the Kafka cluster. The client will make use of all servers irrespective of which servers are specified here for bootstrapping; this list only impacts the initial hosts used to discover the full set of servers. This list should be in the form host1:port1,host2:port2,.... Since these servers are just used for the initial connection to discover the full cluster membership (which may change dynamically), this list need not contain the full set of servers (you may want more than one, though, in case a server is down).

CONTROL_CENTER_REPLICATION_FACTOR

Replication factor for Control Center topics. We recommend setting this to 3 in a production environment.

Optional Settings¶

CONTROL_CENTER_CONNECT_CLUSTER

To enable Control Center to interact with a Kafka Connect cluster, set this parameter to the REST endpoint URL for the Kafka Connect cluster.

Confluent Enterprise Replicator ¶

Confluent Kafka Replicator is a Kafka connector and runs on a Kafka Connect cluster. The image uses variables prefixed with CONNECT_ with an underscore (_) separating each word instead of periods. As an example, to set the required properties like bootstrap.servers, the topic names for config, offsets and status as well the key or value converter, run the following command:

docker run -d \
  --name=cp-enterprise-replicator \
  --net=host \
  -e CONNECT_BOOTSTRAP_SERVERS=localhost:29092 \
  -e CONNECT_REST_PORT=28082 \
  -e CONNECT_GROUP_ID="quickstart" \
  -e CONNECT_CONFIG_STORAGE_TOPIC="quickstart-config" \
  -e CONNECT_OFFSET_STORAGE_TOPIC="quickstart-offsets" \
  -e CONNECT_STATUS_STORAGE_TOPIC="quickstart-status" \
  -e CONNECT_KEY_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_VALUE_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_INTERNAL_KEY_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_INTERNAL_VALUE_CONVERTER="org.apache.kafka.connect.json.JsonConverter" \
  -e CONNECT_REST_ADVERTISED_HOST_NAME="localhost" \
  confluentinc/cp-kafka-connect:4.0.0

The following example shows how to create a Confluent Kafka Replicator connector which replicates topic “confluent” from source Kafka cluster (src) to a destination Kafka cluster (dest).

curl -X POST \
     -H "Content-Type: application/json" \
     --data '{
        "name": "confluent-src-to-dest",
        "config": {
          "connector.class":"io.confluent.connect.replicator.ReplicatorSourceConnector",
          "key.converter": "io.confluent.connect.replicator.util.ByteArrayConverter",
          "value.converter": "io.confluent.connect.replicator.util.ByteArrayConverter",
          "src.zookeeper.connect": "zookeeper-src:2181",
          "src.kafka.bootstrap.servers": "kafka-src:9082",
          "dest.zookeeper.connect": "zookeeper-dest:2181",
          "topic.whitelist": "confluent",
          "topic.rename.format": "${topic}.replica"}}'  \
            http://localhost:28082/connectors

Required Settings¶

The following settings must be passed to run the Kafka Connect Docker image:

CONNECT_BOOTSTRAP_SERVERS

A unique string that identifies the Connect cluster group this worker belongs to.

CONNECT_GROUP_ID

A unique string that identifies the Connect cluster group this worker belongs to.

CONNECT_CONFIG_STORAGE_TOPIC

The name of the topic in which to store connector and task configuration data. This must be the same for all workers with the same group.id

CONNECT_OFFSET_STORAGE_TOPIC

The name of the topic in which to store offset data for connectors. This must be the same for all workers with the same group.id

CONNECT_STATUS_STORAGE_TOPIC

The name of the topic in which to store state for connectors. This must be the same for all workers with the same group.id

CONNECT_KEY_CONVERTER

Converter class for keys. This controls the format of the data that will be written to Kafka for source connectors or read from Kafka for sink connectors.

CONNECT_VALUE_CONVERTER

Converter class for values. This controls the format of the data that will be written to Kafka for source connectors or read from Kafka for sink connectors.

CONNECT_INTERNAL_KEY_CONVERTER

Converter class for internal keys that implements the Converter interface.

CONNECT_INTERNAL_VALUE_CONVERTER

Converter class for internal values that implements the Converter interface.

CONNECT_REST_ADVERTISED_HOST_NAME

Advertised host name is required for starting up the Docker image because it is important to think through how other clients are going to connect to Connect REST API. In a Docker environment, you will need to make sure that your clients can connect to Connect and other services. Advertised host name is how Connect gives out a host name that can be reached by the client.

Optional Settings¶

All other settings for Connect like security, monitoring interceptors, producer and consumer overrides can be passed to the Docker images as environment variables. The names of these environment variables are derived by replacing . with _, converting the resulting string to uppercase and prefixing it with CONNECT_. For example, if you need to set ssl.key.password, the environment variable name would be CONNECT_SSL_KEY_PASSWORD.

The image will then convert these environment variables to corresponding Connect config variables.