Introduction

Kafka Connect allows you to continuously ingest data from external systems into Kafka via connect sources and write data from Kafka to external system via connect sinks.

Various plugins are available for a variety of different data sources and data sinks.

Confluent Plugins

Camel Plugins

Debezium Plugins

Single Message Transformations (SMTs) are applied to messages as they flow through Connect.

Connect Demo

Postgres JDBC Sink Connector (Confluent)

Prepare the connector image

Download the JDBC sink connector from JDBC Connector (Source and Sink)

Create a docker file for this connector

Docker

FROM quay.io/strimzi/kafka:0.32.0-kafka-3.3.1
USER root:root
RUN mkdir -p /opt/kafka/plugins/jdbc
COPY ./confluentinc-kafka-connect-jdbc-10.6.0.zip /opt/kafka/plugins/j
dbc/
RUN unzip /opt/kafka/plugins/jdbc/confluentinc-kafka-connect-jdbc-10.6.0.zip -d /opt/kafka/plugins/jdbc
RUN rm /opt/kafka/plugins/jdbc/confluentinc-kafka-connect-jdbc-10.6.0.zip
USER 1001

Build the image and copy it to your docker repository

docker build -f Dockerfile . t ktimoney/strimzi connector

Prepare the postges database

Setup a new username/password and schema for Kafka connect to write to

Postgres

    SELECT 'CREATE DATABASE kafka'
    WHERE NOT EXISTS (SELECT FROM pg_database WHERE datname = 'kafka')\gexec
    DO $$
    BEGIN
      IF NOT EXISTS (SELECT FROM pg_user WHERE  usename = 'kafka') THEN
         CREATE USER kafka WITH PASSWORD 'kafka';
         GRANT ALL PRIVILEGES ON DATABASE kafka TO kafka;
      END IF;
    END
    $$;

Prepare the message producer

In this example we are going to deserialize the kafka key and value to json, for this to work we need to include the json schema with the message.

They will take the following format:

Schema & Payload

{
  "schema": {
    "type": "struct",
    "optional": false,
    "version": 1,
    "fields": [
      {
        "field": "id",
        "type": "string",
        "optional": true
      }
    ]
  },
  "payload": {
    "id": 1
  }
}

The above structure can be represented in golang using the following structs:

Go Structs

type SchemaPayload[T any] struct {
        Schema Schema `json:"schema"`
        Payload T `json:"payload"`
}

type Schema struct {
        Type string `json:"type"`
        Optional bool  `json:"optional"`
        Version int `json:"version"`
        Fields []Fields `json:"fields"`
}

type Fields struct {
        Field string  `json:"field"`
        Type string `json:"type"`
        Optional bool  `json:"optional"`
}

type KeyPayload struct {
      Id       string      `json:"id"`
}

type ValuePayload struct {
      Text       string      `json:"text"`
}

We can use the same SchemaPayload struct for both keys and values, we only need to provide the Payload struct type we initialize it with.

Create the KafkaConnect object

KafkaConnect

apiVersion: kafka.strimzi.io/v1beta1
kind: KafkaUser
metadata:
  name: connect
  namespace: kafka
  labels:
    strimzi.io/cluster: my-cluster
spec:
  authentication:
    type: tls
---
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
  name: my-connect-cluster
  namespace: kafka
  annotations:
  # use-connector-resources configures this KafkaConnect
  # to use KafkaConnector resources to avoid
  # needing to call the Connect REST API directly
    strimzi.io/use-connector-resources: "true"
spec:
  image: ktimoney/strimzi-connect
  replicas: 1
  bootstrapServers: my-cluster-kafka-bootstrap:9093
  tls:
    trustedCertificates:
      - secretName: my-cluster-cluster-ca-cert
        certificate: ca.crt
      - secretName: my-cluster-clients-ca-cert
        certificate: ca.crt
  authentication:
    type: tls
    certificateAndKey:
      secretName: connect
      certificate: user.crt
      key: user.key
  config:
    group.id: connect-cluster
    offset.storage.topic: connect-cluster-offsets
    config.storage.topic: connect-cluster-configs
    status.storage.topic: connect-cluster-status
    # -1 means it will use the default replication factor configured in the broker
    config.storage.replication.factor: -1
    offset.storage.replication.factor: -1
    status.storage.replication.factor: -1

We are connecting to the 9093 tls port so we include required certificates in our configuration.

The image tag references the image we built earlier : ktimoney/strimzi-connect

Create the Sink connector

KafkaConnector

apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnector
metadata:
  name: my-sink-connector
  namespace: kafka
  labels:
    strimzi.io/cluster: my-connect-cluster
spec:
  class: io.confluent.connect.jdbc.JdbcSinkConnector
  tasksMax: 2
  config:
    topics: my-topic
    connection.url: jdbc:postgresql://postgres.default:5432/kafka
    connection.user: kafka
    connection.password: kafka
    key.converter: org.apache.kafka.connect.json.JsonConverter 
    value.converter: org.apache.kafka.connect.json.JsonConverter
    key.converter.schemas.enable: true
    value.converter.schemas.enable: true
    pk.mode: record_key
    pk.fields: id
    auto.create: true
    delete.enabled: true

The KafkaConnector is configured to use the "io.confluent.connect.jdbc.JdbcSinkConnector" class.

The converters are set to "org.apache.kafka.connect.json.JsonConverter"

The pk.mode is set to "record_key" and pk.fields is set to "id", this means it will use the id field from the key as the primary key.

The database connection details are also included.

Influxdb Sink Connector (Apache Camel)

Prepare the connector image

Download the Influxdb connector: wget https://repo.maven.apache.org/maven2/org/apache/camel/kafkaconnector/camel-influxdb-kafka-connector/0.8.0/camel-influxdb-kafka-connector-0.8.0-package.tar.gz

Create a docker file for this connector

Docker

FROM quay.io/strimzi/kafka:0.22.1-kafka-2.7.0
USER root:root 
RUN mkdir -p /opt/kafka/plugins/camel
COPY ./camel-influxdb-kafka-connector-0.8.0-package.tar.gz /opt/kafka/plugins/camel/
RUN tar -xvzf /opt/kafka/plugins/camel/camel-influxdb-kafka-connector-0.8.0-package.tar.gz --directory /opt/kafka/plugins/camel
RUN rm /opt/kafka/plugins/camel/camel-influxdb-kafka-connector-0.8.0-package.tar.gz 
USER 1001

Build the image and copy it to your docker repository

docker build -f Dockerfile . t ktimoney/strimzi_influxdb_connector

Prepare the message producer

In this example we are going to deserialize the value to json, we can do this without using a schema.

They will take the following format:

Schema & Payload

{
"camelInfluxDB.MeasurementName":"disk_space"
"time":"2023-01-23T13:03:31Z"
"host":"localhost"
"region":"IE-region"
"used":"19%"
"free":"81%"
}

The important thing to note here is that one of the "keys" has to be "camelInfluxDB.MeasurementName", this is required by the camel transformer when converting into an influxdb point object.

The above structure can be represented in golang using the following struct:

Go Structs

type InfluxPayload struct {
        Measurement string `json:"camelInfluxDB.MeasurementName"`
        Time        string `json:"time"`
        Host   string `json:"host"`
        Region string `json:"region"`
        Used   string `json:"used"`
        Free   string `json:"free"`
}

Create the KafkaConnect object

KafkaConnect

apiVersion: kafka.strimzi.io/v1beta1
kind: KafkaUser
metadata:
  name: connect
  namespace: kafka
  labels:
    strimzi.io/cluster: my-cluster
spec:
  authentication:
    type: tls
---
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
  name: my-connect-cluster
  namespace: kafka
  annotations:
  # use-connector-resources configures this KafkaConnect
  # to use KafkaConnector resources to avoid
  # needing to call the Connect REST API directly
    strimzi.io/use-connector-resources: "true"
spec:   
  image: ktimoney/strimzi-influxdb-connect
  replicas: 1
  bootstrapServers: my-cluster-kafka-bootstrap:9093
  tls:
    trustedCertificates:
      - secretName: my-cluster-cluster-ca-cert
        certificate: ca.crt
      - secretName: my-cluster-clients-ca-cert
        certificate: ca.crt
  authentication:
    type: tls
    certificateAndKey:
      secretName: connect
      certificate: user.crt
      key: user.key
  config:
    group.id: connect-cluster
    offset.storage.topic: connect-cluster-offsets
    config.storage.topic: connect-cluster-configs
    status.storage.topic: connect-cluster-status
    config.storage.replication.factor: 1
    offset.storage.replication.factor: 1
    status.storage.replication.factor: 1

We are connecting to the 9093 tls port so we include required certificates in our configuration.

The image tag references the image we built earlier : ktimoney/strimzi-influx-connect

Create the Sink connector

KafkaConnector

apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnector
metadata:
  name: my-sink-connector
  namespace: kafka
  labels:
    strimzi.io/cluster: my-connect-cluster
spec:   class: org.apache.camel.kafkaconnector.influxdb.CamelInfluxdbSinkConnector
  tasksMax: 1
  config:
    topics: my-topic
    errors.deadletterqueue.topic.name: my-topic-dl
    errors.deadletterqueue.topic.replication.factor: 1
    key.converter: org.apache.kafka.connect.storage.StringConverter
    value.converter: org.apache.kafka.connect.json.JsonConverter 
    key.converter.schemas.enable: false
    value.converter.schemas.enable: false
    key.ignore: true
    auto.create: true
    camel.beans.influx: "#class:org.influxdb.InfluxDBFactory#connect('http://influxdb.default:8086', 'influxdb', 'influxdb')"
    camel.sink.path.connectionBean: influx
    camel.sink.endpoint.databaseName: ts_host_metrics
    camel.sink.endpoint.operation: insert
    camel.sink.endpoint.retentionPolicy: autogen

The KafkaConnector is configured to use the "org.apache.camel.kafkaconnector.influxdb.CamelInfluxdbSinkConnector" class.

The value converter is set to "org.apache.kafka.connect.json.JsonConverter".

The camel connector comes with a in built TypeConverter called CamelInfluxDbConverters which will read the json produced as a Map<String, Object> map) and return an Influxdb Point object.

The database connection details are specified in camel.beans.influx.

This sets up a connection bean to be used by camel.sink.path.connectionBean.

Note: It is important to wrap the camel.beans.influx parameter in quotes otherwise it will be treated as a comment.

The yaml for the influxdb pod is available here: influxdb.yaml

When the connector is running it will produce records to influxdb like the following:

InfluxDB query

{'pretty': 'true', 'db': 'ts_host_metrics', 'q': 'SELECT "region", "host", "free", "used" FROM "disk_space" WHERE "host"=\'localhost\''}
{
    "results": [
        {
            "statement_id": 0,
            "series": [
                {
                    "name": "disk_space",
                    "columns": [
                        "time",
                        "region",
                        "host",
                        "free",
                        "used"
                    ],
                    "values": [
                        [
                            "2023-01-23T13:03:32.1246436Z",
                            "IE-region",
                            "localhost",
                            "81%",
                            "19%"
                        ]
                    ]
                }
            ]
        }
    ]
}

Minio Sink Connector (Apache Camel)

Prepare the connector image

Download the minio sink connector: wget https://repo.maven.apache.org/maven2/org/apache/camel/kafkaconnector/camel-file-kafka-connector/3.20.0/camel-file-kafka-connector-3.20.0-package.tar.gz

We need to create some custom classes:

CamelMinioConverters.java - this will allow us to deserialize to json and convert the hashmap to an inputstream

CamelMinioConverters

package com.custom.convert.minio;

import com.fasterxml.jackson.core.JsonProcessingException;
import com.fasterxml.jackson.databind.ObjectMapper;
import java.io.ByteArrayInputStream;
import java.io.InputStream;
import java.util.Map;
import org.apache.camel.Converter;

@Converter(generateLoader = true)
public final class CamelMinioConverters {

    private CamelMinioConverters() {
    }

    @SuppressWarnings("deprecation")
    @Converter
    public static InputStream fromMapToInputStream(Map<String, Object> map) {

      String json = "{}";
      try {
        json = new ObjectMapper().writeValueAsString(map);
      } catch (JsonProcessingException e) {
        e.printStackTrace();
      }
     return new ByteArrayInputStream(json.getBytes());
    }

}

StringAggregator.java - this will allow us to aggregate the messages and set the file name

StringAggregator

package com.custom.aggregate.minio;

import java.text.SimpleDateFormat;
import java.util.Date;
import java.util.HashMap;
import java.util.Map;
import org.apache.camel.AggregationStrategy;
import org.apache.camel.Exchange;
import org.apache.camel.Message;

public class CustomStringAggregator implements AggregationStrategy {

    //@Override
    public Exchange aggregate(Exchange oldExchange, Exchange newExchange) {
      Message newIn = newExchange.getIn();
      Map<String, String> keyMap = (HashMap) newIn.getHeaders().get("camel.kafka.connector.record.key");
      String key = keyMap.get("id");
      SimpleDateFormat sdf = new SimpleDateFormat("ddMMyy-hhmmss-SSS");              
      String fileName = "ExchangeKafka-" + key + "-" + sdf.format( new Date() ) + ".json";
      newIn.setHeader("file", fileName);               

     // lets append the old body to the new body
        if (oldExchange == null) {
            return newExchange;
        }

        String body = oldExchange.getIn().getBody(String.class);
        if (body != null) {
            String newBody = newIn.getBody(String.class);
            if (newBody != null) {
                body += System.lineSeparator() + newBody;
            }

            newIn.setBody(body);
        }
        return newExchange;
    }
}

We can then create a custom jar for our code and copy it into the docker container.

Create a docker file for this connector

Docker

FROM quay.io/strimzi/kafka:0.32.0-kafka-3.3.1
USER root:root
RUN mkdir -p /opt/kafka/plugins/camel
COPY ./camel-minio-sink-kafka-connector-3.20.0-package.tar.gz /opt/kafka/plugins/camel/
RUN tar -xvzf /opt/kafka/plugins/camel/camel-minio-sink-kafka-connector-3.20.0-package.tar.gz --directory /opt/kafka/plugins/camel
COPY ./custom-converter-1.0.0.jar /opt/kafka/plugins/camel/camel-minio-sink-kafka-connector/
RUN rm /opt/kafka/plugins/camel/camel-minio-sink-kafka-connector-3.20.0-package.tar.gzRUN rm -rf /opt/kafka/plugins/camel/docs
USER 1001

Build the image and copy it to your docker repository

docker build -f Dockerfile . t ktimoney/strimzi_minio_connector

Prepare the message producer

In this example we are going to deserialize the value to a string. (Same producer as the influxdb example)

They will take the following format:

Schema & Payload

{
"camelInfluxDB.MeasurementName":"disk_space"
"time":"2023-01-23T13:03:31Z"
"host":"localhost"
"region":"IE-region"
"used":"19%"
"free":"81%"
}

With the type converter n place we can also deserialize as a json object.

They key will be a json that loos like this: {"id":"119"}

Create the KafkaConnect object

KafkaConnect

apiVersion: kafka.strimzi.io/v1beta1
kind: KafkaUser
metadata:
  name: connect
  namespace: kafka
  labels:
    strimzi.io/cluster: my-cluster
spec:
  authentication:
    type: tls
---
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
  name: my-connect-cluster
  namespace: kafka
  annotations:
  # use-connector-resources configures this KafkaConnect
  # to use KafkaConnector resources to avoid
  # needing to call the Connect REST API directly
    strimzi.io/use-connector-resources: "true"
spec:   
  image: ktimoney/strimzi-minio-connect
  replicas: 1
  bootstrapServers: my-cluster-kafka-bootstrap:9093
  tls:
    trustedCertificates:
      - secretName: my-cluster-cluster-ca-cert
        certificate: ca.crt
      - secretName: my-cluster-clients-ca-cert
        certificate: ca.crt
  authentication:
    type: tls
    certificateAndKey:
      secretName: connect
      certificate: user.crt
      key: user.key
  config:
    group.id: connect-cluster      
    offset.storage.topic: connect-cluster-offsets
    config.storage.topic: connect-cluster-configs
    status.storage.topic: connect-cluster-status
    # -1 means it will use the default replication factor configured in the broker
    config.storage.replication.factor: -1
    offset.storage.replication.factor: -1
    status.storage.replication.factor: -1

We are connecting to the 9093 tls port so we include required certificates in our configuration.

The image tag references the image we built earlier : ktimoney/strimzi-minio-connect

Create the Sink connector

KafkaConnector

apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnector
metadata:
  name: my-sink-connector
  namespace: kafka
  labels:
    strimzi.io/cluster: my-connect-cluster
spec:   
  class: org.apache.camel.kafkaconnector.miniosink.CamelMiniosinkSinkConnector   
  tasksMax: 1   
  config:
    topics: my-topic
    key.converter: org.apache.kafka.connect.json.JsonConverter
    value.converter: org.apache.kafka.connect.storage.StringConverter
    key.converter.schemas.enable: false
    value.converter.schemas.enable: false         
    camel.beans.aggregate: "#class:com.custom.aggregate.minio.CustomStringAggregator"
    camel.aggregation.size: 1
    camel.aggregation.timeout: 20000
    camel.aggregation.disable: false     
    camel.kamelet.minio-sink.bucketName: camel
    camel.kamelet.minio-sink.accessKey: ybK7RleFUkdDeYBf
    camel.kamelet.minio-sink.secretKey: X0Y4zK84bwdefRTljrnPzOb1l0A6OQj2
    camel.kamelet.minio-sink.endpoint: http://minio.default:9000
    camel.kamelet.minio-sink.autoCreateBucket: true

The KafkaConnector is configured to use the "org.apache.camel.kafkaconnector.miniosink.CamelMiniosinkSinkConnector class.

The value converter is set to "org.apache.kafka.connect.storage.StringConverter".

The minio connection details are specified in camel.kamelet.minio-sink parameters.

The aggregation size is set to 1 so it won't aggregate any of the files

This timeout is also specified so it will only aggregate records arriving within this time out value (20 seconds).

These values can be adjusted depending on your requirements.

The main purpose of the aggregator in this example is for setting the file name, this is done by setting a header with a key of "file".

The yaml for the minio pod is available here: minio.yaml

When the connector is running it copies the kafka records into minio as files:

Here's a screen shot where we use the aggregator to set the file name:

File Sink Connector (Apache Camel)

Prepare the connector image

Download the file connector: wget https://repo.maven.apache.org/maven2/org/apache/camel/kafkaconnector/camel-file-kafka-connector/3.20.0/camel-file-kafka-connector-3.20.0-package.tar.gz

Create a docker file for this connector

Docker

FROM quay.io/strimzi/kafka:0.32.0-kafka-3.3.1
USER root:root
RUN mkdir -p /opt/kafka/plugins/camel
COPY ./camel-file-kafka-connector-3.20.0-package.tar.gz /opt/kafka/plugins/camel/
RUN tar -xvzf /opt/kafka/plugins/camel/camel-file-kafka-connector-3.20.0-package.tar.gz --directory /opt/kafka/plugins/camel
RUN rm /opt/kafka/plugins/camel/camel-file-kafka-connector-3.20.0-package.tar.gz
RUN rm -rf /opt/kafka/plugins/camel/docs
USER 1001

Build the image and copy it to your docker repository

docker build -f Dockerfile . t ktimoney/strimzi_file_connector

Prepare the message producer

In this example we are going to deserialize the value to a string.

The value will take the following format:

Schema & Payload

{
"camelInfluxDB.MeasurementName":"disk_space"
"time":"2023-01-23T13:03:31Z"
"host":"localhost"
"region":"IE-region"
"used":"19%"
"free":"81%"
}

Due to some limitation with Strimzi KafkaConnect we will need to use a mutating webhook to attach a persistent volume to our connector pod.

The go code for this is available here: kafka-webhook.go

We need to create a docker image for this program

Webhook docker

FROM golang:latest
RUN mkdir /app
COPY ./kafka-webhook /app
RUN chmod +x /app/kafka-webhook
WORKDIR /app
ENTRYPOINT ["/app/kafka-webhook"]

docker build -f Dockerfile . -t ktimoney/kafka-webhook

Mutating webhooks require https so we also need t create some certificates.

Make sure cfssl is installed on you system, then run the following commands.

cfssl gencert -initca ./ca-csr.json | cfssljson -bare ca

cfssl gencert \
-ca=ca.pem \
-ca-key=ca-key.pem \
-config=ca-config.json \
-hostname="kafka-webhook,kafka-webhook.kafka.svc.cluster.local,kafka-webhook.kafka.svc,localhost,127.0.0.1" \
-profile=default \
ca-csr.json | cfssljson -bare webhook-cert

Run the following commands to create the values you need to use in the yaml file:

TLS_CRT=$(cat webhook-cert.pem | base64 | tr -d '\n')

TLS_KEY=$(cat webhook-cert-key.pem | base64 | tr -d '\n')

CA_BUNDLE="$(openssl base64 -A <"ca.pem")"

Replace the values in : kafka-webhook.yaml

Note: You can also configure the hostPath and the containerPath as part of the deployment:

Container command args

      containers:
      - name: kafka-webhook
        image: ktimoney/kafka-webhook
        imagePullPolicy: IfNotPresent
        command: ["/app/kafka-webhook"]
        args: [
                "-port", "8443",
                "-tlsCertFile", "/certs/tls.crt",
                "-tlsKeyFile", "/certs/tls.key",
                "-hostPath", "/var/strimzi/files",
                "-containerPath", "/opt/kafka/data"
              ]

Run: kubectl create -f kafka webhook.yaml to start the pod.

Create the KafkaConnect object

KafkaConnect

apiVersion: kafka.strimzi.io/v1beta1
kind: KafkaUser
metadata:
  name: connect
  namespace: kafka
  labels:
    strimzi.io/cluster: my-cluster
spec:
  authentication:
    type: tls
---
apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnect
metadata:
  name: my-connect-cluster
  namespace: kafka
  annotations:
  # use-connector-resources configures this KafkaConnect
  # to use KafkaConnector resources to avoid
  # needing to call the Connect REST API directly
    strimzi.io/use-connector-resources: "true"
spec:   
  image: ktimoney/strimzi-file-connect
  replicas: 1
  bootstrapServers: my-cluster-kafka-bootstrap:9093
  tls:
    trustedCertificates:
      - secretName: my-cluster-cluster-ca-cert
        certificate: ca.crt
      - secretName: my-cluster-clients-ca-cert
        certificate: ca.crt
  authentication:
    type: tls
    certificateAndKey:
      secretName: connect
      certificate: user.crt
      key: user.key
  config:
    group.id: connect-cluster
    offset.storage.topic: connect-cluster-offsets
    config.storage.topic: connect-cluster-configs
    status.storage.topic: connect-cluster-status
    config.storage.replication.factor: 1
    offset.storage.replication.factor: 1
    status.storage.replication.factor: 1

We are connecting to the 9093 tls port so we include required certificates in our configuration.

The image tag references the image we built earlier : ktimoney/strimzi-file-connect

Create the Sink connector

KafkaConnector

apiVersion: kafka.strimzi.io/v1beta2
kind: KafkaConnector
metadata:
  name: my-sink-connector
  namespace: kafka
  labels:
    strimzi.io/cluster: my-connect-cluster
spec:   
  class: org.apache.camel.kafkaconnector.file.CamelFileSinkConnector
  tasksMax: 1
  config:
    topics: my-topic 
    value.converter: org.apache.kafka.connect.storage.StringConverter
    key.converter: org.apache.kafka.connect.json.JsonConverter
    key.converter.schemas.enable: false
    value.converter.schemas.enable: false
    camel.sink.endpoint.fileName: mydata-${date:now:ddMMyy-hhmmss-SSS}.txt
    camel.sink.path.directoryName: /opt/kafka/data
    camel.sink.endpoint.fileExist: Append
    camel.sink.endpoint.autoCreate: true
    camel.aggregation.disable: true

The KafkaConnector is configured to use the "org.apache.camel.kafkaconnector.file.CamelFileSinkConnector" class.

The value converter is set to "org.apache.kafka.connect.storage.StringConverter".

The connector filename is set using the expression: mydata-${date:now:ddMMyy-hhmmss-SSS}.txt

When the connector starts your kafka records will be copied to the persistent volume as files: