Mounting Docker External Volumes

When working with Docker, you may sometimes need to persist data in the event of a container going down or share data across containers. In order to do so, you can use Docker Volumes. For Confluent Platform, you need to use external volumes for several main use cases:

  1. Data Storage: Apache Kafka® and ZooKeeper will need externally mounted volumes to persist data in the event that a container stops running or is restarted.

  2. Security: When security is configured, the secrets are stored on the host and made available to the containers using mapped volumes.

  3. Configuring Connect with external JARs: Kafka Connect can be configured to use third-party jars by storing them on a volume on the host.

    Note

    If you need to add support for additional use cases for external volumes, see extending the images.

Data volumes for Kafka and ZooKeeper

Kafka uses volumes for log data and ZooKeeper uses volumes for transaction logs. You should separate volumes on the host for these services. You must also ensure that the host directory has read/write permissions for the Docker container user, which has root permissions by default unless you assigned a user using the Docker run command.

Important

When mapping volumes from host, you must use full path (example: /var/lib/kafka/data).

The following example shows how to use Kafka and ZooKeeper with mounted volumes and how to configure volumes if you are running Docker container as non-root user. In this example, the containers run with the user appuser with UID=1000 and GID=1000. In all Confluent Platform images, the containers run with the appuser user.

On the Docker host (VirtualBox VM for example), create the directories:

# Create dirs for Kafka / ZK data.
mkdir -p /vol1/zk-data
mkdir -p /vol2/zk-txn-logs
mkdir -p /vol3/kafka-data

# Make sure the user has the read and write permissions.
chown -R 1000:1000 /vol1/zk-data
chown -R 1000:1000 /vol2/zk-txn-logs
chown -R 1000:1000 /vol3/kafka-data

Then start the containers:

# Run Zookeeper with user 1000 and volumes mapped to host volumes
docker run -d \
  --name=zk-vols \
  --net=host \
  --user=1000 \
  -e ZOOKEEPER_TICK_TIME=2000 \
  -e ZOOKEEPER_CLIENT_PORT=32181 \
  -v /vol1/zk-data:/var/lib/zookeeper/data \
  -v /vol2/zk-txn-logs:/var/lib/zookeeper/log \
  confluentinc/cp-zookeeper:7.3.10

docker run -d \
  --name=kafka-vols \
  --net=host \
  --user=1000 \
  -e KAFKA_BROKER_ID=1 \
  -e KAFKA_ZOOKEEPER_CONNECT=localhost:32181 \
  -e KAFKA_ADVERTISED_LISTENERS=PLAINTEXT://localhost:39092 \
  -e KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR=1 \
  -v /vol3/kafka-data:/var/lib/kafka/data \
  confluentinc/cp-kafka:7.3.10

The data volumes are mounted using the -v flag.

Security: Data volumes for configuring secrets

When security is enabled, the secrets are made available to the containers using volumes. For example, if the host has the secrets (credentials, keytab, certificates, kerberos config, JAAS config) in /vol007/kafka-node-1-secrets, we can configure Kafka as follows to use the secrets:

docker run -d \
  --name=kafka-sasl-ssl-1 \
  --net=host \
  -e KAFKA_BROKER_ID=1 \
  -e KAFKA_ZOOKEEPER_CONNECT=localhost:22181,localhost:32181,localhost:42181/saslssl \
  -e KAFKA_ADVERTISED_LISTENERS=SASL_SSL://localhost:39094 \
  -e KAFKA_SSL_KEYSTORE_FILENAME=kafka.broker3.keystore.jks \
  -e KAFKA_SSL_KEYSTORE_CREDENTIALS=broker3_keystore_creds \
  -e KAFKA_SSL_KEY_CREDENTIALS=broker3_sslkey_creds \
  -e KAFKA_SSL_TRUSTSTORE_FILENAME=kafka.broker3.truststore.jks \
  -e KAFKA_SSL_TRUSTSTORE_CREDENTIALS=broker3_truststore_creds \
  -e KAFKA_SECURITY_INTER_BROKER_PROTOCOL=SASL_SSL \
  -e KAFKA_SASL_MECHANISM_INTER_BROKER_PROTOCOL=GSSAPI \
  -e KAFKA_SASL_ENABLED_MECHANISMS=GSSAPI \
  -e KAFKA_SASL_KERBEROS_SERVICE_NAME=kafka \
  -e KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR=1 \
  -e KAFKA_OPTS=-Djava.security.auth.login.config=/etc/kafka/secrets/host_broker3_jaas.conf -Djava.security.krb5.conf=/etc/kafka/secrets/host_krb.conf \
  -v /vol007/kafka-node-1-secrets:/etc/kafka/secrets \
  confluentinc/cp-kafka:latest

In the example above, the location of the data volumes is specified by setting -v /vol007/kafka-node-1-secrets:/etc/kafka/secrets. You then specify how they are to be used by setting:

-e KAFKA_OPTS=-Djava.security.auth.login.config=/etc/kafka/secrets/host_broker3_jaas.conf -Djava.security.krb5.conf=/etc/kafka/secrets/host_krb.conf

Configuring Connect with external JARs

Kafka Connect can be configured to use third-party jars by storing them on a volume on the host and mapping the volume to /etc/kafka-connect/jars on the container.

At the host (Virtualbox VM for example), download the MySQL driver:

# Create a dir for jars and download the mysql jdbc driver into the directories
mkdir -p /vol42/kafka-connect/jars

# get the driver and store the jar in the dir
curl -k -SL "https://dev.mysql.com/get/Downloads/Connector-J/mysql-connector-java-5.1.39.tar.gz" | tar -xzf - -C /vol42/kafka-connect/jars --strip-components=1 mysql-connector-java-5.1.39/mysql-connector-java-5.1.39-bin.jar

Then start Kafka Connect mounting the download directory as /etc/kafka-connect/jars:

docker run -d \
  --name=connect-host-json \
  --net=host \
  -e CONNECT_BOOTSTRAP_SERVERS=localhost:39092 \
  -e CONNECT_REST_PORT=28082 \
  -e CONNECT_GROUP_ID="default" \
  -e CONNECT_CONFIG_STORAGE_TOPIC="default.config" \
  -e CONNECT_OFFSET_STORAGE_TOPIC="default.offsets" \
  -e CONNECT_STATUS_STORAGE_TOPIC="default.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" \
  -e CONNECT_PLUGIN_PATH=/usr/share/java,/etc/kafka-connect/jars \
  -e KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR=1 \
  -v /vol42/kafka-connect/jars:/etc/kafka-connect/jars \
  confluentinc/cp-kafka-connect:latest