Kafka GUI for Apache Kafka to manage topics, topics data, consumers group, schema registry, connect and more...

• Features
• Quick Preview
• Installation
  • Docker
  • Stand Alone
  • Kubernetes using Helm
• Configuration
  • JVM.options file
  • Kafka cluster
  • AKHQ
  • Security
  • Server
  • Micronaut
• Api
• Monitoring Endpoint
• Development Environment
• Schema references
• Who's using AKHQ

• General
  • Works with modern Kafka cluster (1.0+)
  • Connection on standard or ssl, sasl cluster
  • Multi cluster
• Topics
  • List
  • Configurations view
  • Partitions view
  • ACLS view
  • Consumers groups assignments view
  • Node leader & assignments view
  • Create a topic
  • Configure a topic
  • Delete a topic
• Browse Topic datas
  • View data, offset, key, timestamp & headers
  • Automatic deserializarion of avro message encoded with schema registry
  • Configurations view
  • Logs view
  • Delete a record
  • Empty a Topic (Delete all the record from one topic)
  • Sort view
  • Filter per partitions
  • Filter with a starting time
  • Filter data with a search string
• Consumer Groups (only with kafka internal storage, not with old Zookeeper)
  • List with lag, topics assignments
  • Partitions view & lag
  • ACLS view
  • Node leader & assignments view
  • Display active and pending consumers groups
  • Delete a consumer group
  • Update consumer group offsets to start / end / timestamp
• Schema Registry
  • List schema
  • Create / Update / Delete a schema
  • View and delete individual schema version
• Connect
  • List connect definition
  • Create / Update / Delete a definition
  • Pause / Resume / Restart a definition or a task
• Nodes
  • List
  • Configurations view
  • Logs view
  • Configure a node
• ACLS
  • List principals
  • List principals topic & group acls
• Authentification and Roles
  • Read only mode
  • BasicHttp with roles per user
  • User groups configuration
  • Filter topics with regexp for current groups
  • Ldap configuration to match AKHQ groups/roles

Since this is a major rework, the new UI can have some issues, so please report any issue, thanks!

• Download docker-compose.yml file
• run docker-compose pull to be sure to have the last version of AKHQ
• run docker-compose up
• go to https://localhost:8080

It will start a Kafka node, a Zookeeper node, a Schema Registry, a Connect, fill with some sample data, start a consumer group and a kafka stream & start AKHQ.

First you need a configuration files in order to configure AKHQ connections to Kafka Brokers.

docker run -d \
    -p 8080:8080 \
    -v /tmp/application.yml:/app/application.yml \
    tchiotludo/akhq

• With -v /tmp/application.yml must be an absolute path to configuration file
• Go to https://localhost:8080

• Install Java 11
• Download the latest jar on release page
• Create an configuration files
• Launch the application with java -Dmicronaut.config.files=/path/to/application.yml -jar akhq.jar
• Go to https://localhost:8080

• Add the AKHQ helm charts repository:

helm repo add akhq https://akhq.io/

• Install or upgrade

helm upgrade --install akhq akhq/akhq

• Chart version >=0.1.1 requires Kubernetes version >=1.14
• Chart version 0.1.0 works on previous Kubernetes versions

helm install akhq akhq/akhq --version 0.1.0

• Clone the repository:

git clone https://github.com/tchiotludo/akhq && cd akhq/deploy/helm/akhq

• Update helm values located in deploy/helm/values.yaml
  • configuration values will contains all related configuration that you can find in application.example.yml and will be store in a ConfigMap
  • secrets values will contains all sensitive configurations (with credentials) that you can find in application.example.yml and will be store in Secret
  • Both values will be merged at startup
• Apply the chart:

helm install --name=akhq-release-name  .

Configuration file can by default be provided in either Java properties, YAML, JSON or Groovy files. YML Configuration file example can be found here :application.example.yml

By default, the docker container will allow a custom jvn options setting the environnments vars JAVA_OPTS. For example, if you want to change the default timezome, just add -e "JAVA_OPTS=-Duser.timezone=Europe/Paris"

By default, the docker container will run with a jvm.options file, you can override it with your own with an Environment Variable. With the JVM_OPTS_FILE environment variable, you can override the jvm.options file by passing the path of your file instead.

Override the JVM_OPTS_FILE with docker run:

docker run -d \
    --env JVM_OPTS_FILE={{path-of-your-jvm.options-file}}
    -p 8080:8080 \
    -v /tmp/application.yml:/app/application.yml \
    tchiotludo/akhq

Override the JVM_OPTS_FILE with docker-compose:

version: '3.7'
services:
  akhq:
    image: tchiotludo/akhq-jvm:dev
    environment:
      JVM_OPTS_FILE: /app/jvm.options
    ports:
      - "8080:8080"
    volumes:
      - /tmp/application.yml:/app/application.yml

If you do not override the JVM_OPTS_FILE, the docker container will take the defaults one instead.

• akhq.connections is a key value configuration with :
  • key: must be an url friendly (letter, number, _, -, ... dot are not allowed here) string the identify your cluster (my-cluster-1 and my-cluster-2 is the example above)
  • properties: all the configurations found on Kafka consumer documentation. Most important is bootstrap.servers that is a list of host:port of your Kafka brokers.
  • schema-registry: (optional)
    • url: the schema registry url
    • basic-auth-username: schema registry basic auth username
    • basic-auth-password: schema registry basic auth password
    • properties: all the configurations for registry client, especially ssl configuration
  • connect: (optional list, define each connector as a element of a list)
    • name: connect name
    • url: connect url
    • basic-auth-username: connect basic auth username
    • basic-auth-password: connect basic auth password
    • ssl-trust-store: /app/truststore.jks
    • ssl-trust-store-password: trust-store-password
    • ssl-key-store: /app/truststore.jks
    • ssl-key-store-password: key-store-password

Configuration example for kafka cluster secured by ssl for saas provider like aiven (full https & basic auth):

You need to generate a jks & p12 file from pem, cert files give by saas provider.

openssl pkcs12 -export -inkey service.key -in service.cert -out client.keystore.p12 -name service_key
keytool -import -file ca.pem -alias CA -keystore client.truststore.jks

Configurations will look like this example:

akhq:
  connections:
    ssl-dev:
      properties:
        bootstrap.servers: "{{host}}.aivencloud.com:12835"
        security.protocol: SSL
        ssl.truststore.location: {{path}}/avnadmin.truststore.jks
        ssl.truststore.password: {{password}}
        ssl.keystore.type: "PKCS12"
        ssl.keystore.location: {{path}}/avnadmin.keystore.p12
        ssl.keystore.password: {{password}}
        ssl.key.password: {{password}}
      schema-registry:
        url: "https://{{host}}.aivencloud.com:12838"
        basic-auth-username: avnadmin
        basic-auth-password: {{password}}
        properties: {}
      connect:
        - name: connect-1
          url: "https://{{host}}.aivencloud.com:{{port}}"
          basic-auth-username: avnadmin
          basic-auth-password: {{password}}

• akhq.pagination.page-size number of topics per page (default : 25)

• akhq.topic.internal-regexps is list of regexp to be considered as internal (internal topic can't be deleted or updated)
• akhq.topic.stream-regexps is list of regexp to be considered as internal stream topic

These parameters are the default values used in the topic creation page.

• akhq.topic.replication Default number of replica to use
• akhq.topic.partition Default number of partition

• akhq.topic-data.sort: default sort order (OLDEST, NEWEST) (default: OLDEST)
• akhq.topic-data.size: max record per page (default: 50)
• akhq.topic-data.poll-timeout: The time, in milliseconds, spent waiting in poll if data is not available in the buffer (default: 1000).

• akhq.ui-options.topic.default-view is default list view (ALL, HIDE_INTERNAL, HIDE_INTERNAL_STREAM, HIDE_STREAM) (default: HIDE_INTERNAL)
• akhq.ui-options.topic.skip-consumer-groups hide consumer groups columns on topic list
• akhq.ui-options.topic.skip-last-record hide the last records on topic list

• akhq.ui-options.topic-data.sort: default sort order (OLDEST, NEWEST) (default: OLDEST)

To deserialize topics containing data in Protobuf format, you can set topics mapping: for each topic-regex you can specify descriptor-file-base64 (descriptor file encoded to Base64 format), or you can put descriptor files in descriptors-folder and specify descriptor-file name, also specify corresponding message types for keys and values. If, for example, keys are not in Protobuf format, key-message-type can be omitted, the same for value-message-type. This configuration can be specified for each Kafka cluster.

Example configuration can look like as follows:

akhq:
  connections:
    kafka:
      properties:
        # standard kafka properties
      deserialization:
        protobuf:
          descriptors-folder: "/app/protobuf_desc"
          topics-mapping:
            - topic-regex: "album.*"
              descriptor-file-base64: "Cs4BCgthbGJ1bS5wcm90bxIXY29tLm5ldGNyYWNrZXIucHJvdG9idWYidwoFQWxidW0SFAoFdGl0bGUYASABKAlSBXRpdGxlEhYKBmFydGlzdBgCIAMoCVIGYXJ0aXN0EiEKDHJlbGVhc2VfeWVhchgDIAEoBVILcmVsZWFzZVllYXISHQoKc29uZ190aXRsZRgEIAMoCVIJc29uZ1RpdGxlQiUKF2NvbS5uZXRjcmFja2VyLnByb3RvYnVmQgpBbGJ1bVByb3RvYgZwcm90bzM="
              value-message-type: "Album"
            - topic-regex: "film.*"
              descriptor-file-base64: "CuEBCgpmaWxtLnByb3RvEhRjb20uY29tcGFueS5wcm90b2J1ZiKRAQoERmlsbRISCgRuYW1lGAEgASgJUgRuYW1lEhoKCHByb2R1Y2VyGAIgASgJUghwcm9kdWNlchIhCgxyZWxlYXNlX3llYXIYAyABKAVSC3JlbGVhc2VZZWFyEhoKCGR1cmF0aW9uGAQgASgFUghkdXJhdGlvbhIaCghzdGFycmluZxgFIAMoCVIIc3RhcnJpbmdCIQoUY29tLmNvbXBhbnkucHJvdG9idWZCCUZpbG1Qcm90b2IGcHJvdG8z"
              value-message-type: "Film"
            - topic-regex: "test.*"
              descriptor-file: "other.desc"
              key-message-type: "Row"
              value-message-type: "Envelope"

More examples about Protobuf deserialization can be found in tests. Info about descriptor files generation can be found in test resources.

• akhq.security.default-group: Default group for all the user even unlogged user. By default, the default group is admin and allow you all read / write access on the whole app.

By default, security & roles is enabled by default but anonymous user have full access. You can completely disabled security with micronaut.security.enabled: false.

If you need a read-only application, simply add this to your configuration files :

akhq:
  security:
    default-group: reader

AKHQ uses JWT tokens to perform authentication. Please generate a secret that is at least 256 bits and change the config like this:

micronaut:
  security:
    enabled: true
    token:
      jwt:
        signatures:
          secret:
            generator:
              secret: <Your secret here>

Groups allow you to limit user

Define groups with specific roles for your users

• akhq.security.default-group: Default group for all the user even unlogged user

• akhq.security.groups: Groups map definition

  • key: a uniq key used as name if not specified
    • name: group-name Group identifier
    • roles: Roles list for the group
    • attributes.topics-filter-regexp: Regexp to filter topics available for current group
    • attributes.connects-filter-regexp: Regexp to filter Connect tasks available for current group

3 defaults group are available :

• admin with all right
• reader with only read acces on all AKHQ
• no-roles without any roles, that force user to login

• akhq.security.basic-auth: List user & password with affected roles
  • - username: actual-username: Login of the current user as a yaml key (may be anything email, login, ...)
    • password: Password in sha256 (default) or bcrypt. The password can be converted
      • For default SHA256, with command echo -n "password" | sha256sum or Ansible filter {{ 'password' | hash('sha256') }}
      • For BCrypt, with Ansible filter {{ 'password' | password_hash('blowfish') }}
    • passwordHash: Password hashing algorithm, either SHA256 or BCRYPT
    • groups: Groups for current user

Take care that basic auth will use session store in server memory. If your instance is behind a reverse proxy or a loadbalancer, you will need to forward the session cookie named SESSION and / or use sesssion stickiness

Configure basic-auth connection in AKHQ

akhq.security:
  basic-auth:
    - username: admin
      password: "$2a$<hashed password>"
      passwordHash: BCRYPT
      groups: 
      - admin
    - username: reader
      password: "<SHA-256 hashed password>"
      groups: 
      - reader

Configure how the ldap groups will be matched in AKHQ groups

• akhq.security.ldap.groups: Ldap groups list
  • - name: ldap-group-name: Ldap group name (same name as in ldap)
    • groups: AKHQ group list to be used for current ldap group

Example using online ldap test server

Configure ldap connection in micronaut

micronaut:
  security:
    ldap:
      default:
        enabled: true
        context:
          server: 'ldap:https://ldap.forumsys.com:389'
          managerDn: 'cn=read-only-admin,dc=example,dc=com'
          managerPassword: 'password'
        search:
          base: "dc=example,dc=com"
        groups:
          enabled: true
          base: "dc=example,dc=com"

If you want to enable anonymous auth to your LDAP server you can pass :

managerDn: ''
managerPassword: ''

In Case your LDAP groups do not use the default UID for group membership, you can solve this using

micronaut:
  security:
    ldap:
      default:
        search:
          base: "OU=UserOU,dc=example,dc=com"
          attributes:
            - "cn"
        groups:
          enabled: true
          base: "OU=GroupsOU,dc=example,dc=com"
          filter: "member={0}"

Replace

attributes:
  - "cn"

with your group membership attribute

Debuging ldap connection can be done with

curl -i -X POST -H "Content-Type: application/json" \
       -d '{ "configuredLevel": "TRACE" }' \
       https://localhost:8080/loggers/io.micronaut.configuration.security

Configure AKHQ groups and Ldap groups and users

akhq:
  security:
    groups:
      topic-reader:
        name: topic-reader # Group name
        roles:  # roles for the group
          - topic/read
        attributes:
          # Regexp to filter topic available for group
          topics-filter-regexp: "test\\.reader.*"
          connects-filter-regexp: "^test.*$"
      topic-writer:
        name: topic-writer # Group name
        roles:
          - topic/read
          - topic/insert
          - topic/delete
          - topic/config/update
        attributes:
          topics-filter-regexp: "test.*"
          connects-filter-regexp: "^test.*$"
    ldap:
      groups:
        - name: mathematicians
          groups:
            - topic-reader
        - name: scientists
          groups:
            - topic-reader
            - topic-writer
      users:
        - username: franz
          groups:
            - topic-reader
            - topic-writer

To enable OIDC in the application, you'll first have to enable OIDC in micronaut:

micronaut:
  security:
    oauth2:
      enabled: true
      clients:
        google:
          client-id: "<client-id>"
          client-secret: "<client-secret>"
          openid:
            issuer: "<issuer-url>"

To further tell AKHQ to display OIDC options on the login page and customize claim mapping, configure OIDC in the AKHQ config:

akhq:
  security:
    oidc:
      enabled: true
      providers:
        google:
          label: "Login with Google"
          username-field: preferred_username
          groups-field: roles
          default-group: topic-reader
          groups:
            - name: mathematicians
              groups:
                - topic-reader
            - name: scientists
              groups:
                - topic-reader
                - topic-writer
          users:
            - username: franz
              groups:
                - topic-reader
                - topic-writer

The username field can be any string field, the roles field has to be a JSON array.

• micronaut.server.context-path: if behind a reverse proxy, path to akhq with trailing slash (optional). Example: akhq is behind a reverse proxy with url https://my-server/akhq, set base-path: "/akhq/". Not needed if you're behind a reverse proxy with subdomain https://akhq.my-server/

• akhq.clients-defaults.{{admin|producer|consumer}}.properties: default configuration for admin producer or consumer. All properties from Kafka documentation is available.

Since AKHQ is based on Micronaut, you can customize configurations (server port, ssl, ...) with Micronaut configuration. More information can be found on Micronaut documentation

AKHQ docker image support 3 environment variables to handle configuraiton :

• AKHQ_CONFIGURATION: a string that contains the full configuration in yml that will be written on /app/configuration.yml on container.
• MICRONAUT_APPLICATION_JSON: a string that contains the full configuration in JSON format
• MICRONAUT_CONFIG_FILES: a path to to a configuration file on container. Default path is /app/application.yml

Take care when you mount configuration files to not remove akhq files located on /app. You need to explicitely mount the /app/application.yml and not mount the /app directory. This will remove the AKHQ binnaries and give you this error: /usr/local/bin/docker-entrypoint.sh: 9: exec: ./akhq: not found

volumeMounts:
- mountPath: /app/application.yml
  subPath: application.yml
  name: config
  readOnly: true

An experimental api is available that allow you to fetch all the exposed on AKHQ through api.

Take care that this api is experimental and will change in a future release. Some endpoint expose too many datas and is slow to fetch, and we will remove some properties in a future in order to be fast.

Example: List topic endpoint expose log dir, consumer groups, offsets. Fetching all of theses is slow for now and we will remove these in a future.

You can discover the api endpoint here :

• /api: a RapiDoc webpage that document all the endpoints.
• /swagger/akhq.yml: a full OpenApi specifications files

Several monitoring endpoint is enabled by default. You can disabled it or restrict access only for authenticated users following micronaut configuration below.

• /info Info Endpoint with git status informations.
• /health Health Endpoint
• /loggers Loggers Endpoint
• /metrics Metrics Endpoint
• /prometheus Prometheus Endpoint

You can debug all query duration from AKHQ with this commands

curl -i -X POST -H "Content-Type: application/json" \
       -d '{ "configuredLevel": "TRACE" }' \
       https://localhost:8080/loggers/org.akhq

You can have access to last feature / bug fix with docker dev image automatically build on tag dev

docker pull tchiotludo/akhq:dev

The dev jar is not publish on GitHub, you have 2 solutions to have the dev jar :

Get it from docker image

docker pull tchiotludo/akhq:dev
docker run --rm --name=akhq -it tchiotludo/akhq:dev
docker cp akhq:/app/akhq.jar .

Or build it with a ./gradlew shadowJar, the jar will be located here build/libs/akhq-*.jar

A docker-compose is provide to start a development environnement. Just install docker & docker-compose, clone the repository and issue a simple docker-compose -f docker-compose-dev.yml up to start a dev server. Dev server is a java server & webpack-dev-server with live reload.

The configuration for the dev server is in application.dev.yml.

In case you want to develop for AKHQ on Windows with IntelliJ IDEA without Docker (for any reason) you can follow this brief guide. For the following steps please make sure that you meet this requirements:

• OS: Windows (10)
• Kafka (2.6.0) is downloaded and extracted, the install dir is referred to as $KAFKA_HOME in the latter
• Git is installed and configured
• IntelliJ IDEA (Community Edition 2020.2) with the following plugins installed:
  • Gradle (bundled with IDEA)
  • Lombok

First run a Kafka server locally. Therefore you need to start Zookeper first by opening a CMD and doing:

$KAFKA_HOME\bin\windows\zookeeper-server-start.bat config\zookeper.properties
$KAFKA_HOME\bin\windows\kafka-server-start.bat config\server.properties

A zero-config Kafka server should be up and running locally on your machine now. For further details or troubleshooting see Kafka Getting started guide. In the next step we're going to checkout AKHQ from GitHub:

git clone https://github.com/tchiotludo/akhq.git

Open the checked out directory in IntelliJ IDEA. The current version (0.16.0) of AKHQ is built with Java 11. If you don't have OpenJDK 11 installed already, do the following in IntelliJ IDEA: File > Project Structure... > Platform Settings > SDKs > + > Download JDK... > select a vendor of your choice (but make sure it's version 11), download + install. Make sure that JDK 11 is set under Project Settings > Projekt SDK and language level is Java 11. Now tell Gradle to use Java 11 as well: File > Settings > Plugins > Build, Execution, Deployment > Build Tools > Gradle > Gradle JVM: any JDK 11.

To configure AKHQ for using the Kafka server you set up before, edit application.yml by adding the following under akhq:

akhq:
  connections:
    kafka:
      properties:
        bootstrap.servers: "localhost:9092"

/!\ Do not commit this part of application.yml. A more secure way to configure your local development Kafka server is described in the Micronaut doc, chapter "Application Configuration".

Now you should be able to build the project with Gradle. Therefore go to the Gradle view in IDEA, select Tasks > build > build. If an error occurs saying that any filename is too long: move your project directory to a root directory in your filesystem or as a fix (only for testing purposes) set the argument -x test to skip tests temporarily.

To debug a running AKHQ instance, go to the Gradle tab in IntelliJ IDEA, Tasks > application > right click run and click "Debug(...)". AKHQ should start up and hit the breakpoints you set in your IDE. Happy developing/debugging!

Since Confluent 5.5.0, Avro schemas can now be reused by others schemas through schema references. This feature allows to define a schema once and use it as a record type inside one or more schemas.

When registering new Avro schemas with AKHQ UI, it is now possible to pass a slightly more complex object with a schema and a references field.

To register a new schema without references, no need to change anything:

{
    "name": "Schema1",
    "namespace": "org.akhq",
    "type": "record",
    "fields": [
        {
            "name": "description",
            "type": "string"
        }
    ]
}

To register a new schema with a reference to an already registered schema:

{
    "schema": {
        "name": "Schema2",
        "namespace": "org.akhq",
        "type": "record",
        "fields": [
            {
                "name": "name",
                "type": "string"
            },
            {
                "name": "schema1",
                "type": "Schema1"
            }
        ]
    },
    "references": [
        {
            "name": "Schema1",
            "subject": "SCHEMA_1",
            "version": 1
        }
    ]
}

Documentation on Confluent 5.5 and schema references can be found here.

• Adeo
• Auchan Retail
• Bell
• BMW Group
• Boulanger
• GetYourGuide
• Klarna
• La Redoute
• Leroy Merlin
• NEXT Technologies
• Nuxeo
• Pipedrive
• BARMER
• TVG
• Depop

Many thanks to:

• JetBrains for their free OpenSource license.
• Apache, Apache Kafka, Kafka, and associated open source project names are trademarks of the Apache Software Foundation. AKHQ is not affiliated with, endorsed by, or otherwise associated with the Apache Software.

Apache 2.0 © tchiotludo