Kafdrop – Kafka Web UI  

Kafdrop is a web UI for viewing Kafka topics and browsing consumer groups. The tool displays information such as brokers, topics, partitions, consumers, and lets you view messages.

This project is a reboot of Kafdrop 2.x, dragged kicking and screaming into the world of JDK 11+, Kafka 2.x, Helm and Kubernetes. It's a lightweight application that runs on Spring Boot and is dead-easy to configure, supporting SASL and TLS-secured brokers.

Features

• View Kafka brokers — topic and partition assignments, and controller status
• View topics — partition count, replication status, and custom configuration
• Browse messages — JSON, plain text, Avro and Protobuf encoding
• View consumer groups — per-partition parked offsets, combined and per-partition lag
• Create new topics
• View ACLs
• Support for Azure Event Hubs

Requirements

• Java 11 or newer
• Kafka (version 0.11.0 or newer) or Azure Event Hubs

Optional, additional integration:

• Schema Registry

Getting Started

You can run the Kafdrop JAR directly, via Docker, or in Kubernetes.

Running from JAR

java --add-opens=java.base/sun.nio.ch=ALL-UNNAMED \
    -jar target/kafdrop-<version>.jar \
    --kafka.brokerConnect=<host:port,host:port>,...

If unspecified, kafka.brokerConnect defaults to localhost:9092.

Note: As of Kafdrop 3.10.0, a ZooKeeper connection is no longer required. All necessary cluster information is retrieved via the Kafka admin API.

Open a browser and navigate to http://localhost:9000. The port can be overridden by adding the following config:

--server.port=<port> --management.server.port=<port>

Optionally, configure a schema registry connection with:

--schemaregistry.connect=http://localhost:8081

and if you also require basic auth for your schema registry connection you should add:

--schemaregistry.auth=username:password

Finally, a default message format (e.g. to deserialize Avro messages) can optionally be configured as follows:

--message.format=AVRO

Valid format values are DEFAULT, AVRO, PROTOBUF. This can also be configured at the topic level via dropdown when viewing messages.

Configure Protobuf message type

Option 1: Using Protobuf Descriptor

In case of protobuf message type, the definition of a message could be compiled and transmitted using a descriptor file. Thus, in order for kafdrop to recognize the message, the application will need to access to the descriptor file(s). Kafdrop will allow user to select descriptor and well as specifying name of one of the message type provided by the descriptor at runtime.

To configure a folder with protobuf descriptor file(s) (.desc), follow:

--protobufdesc.directory=/var/protobuf_desc

Option 2 : Using Schema Registry

In case of no protobuf descriptor file being supplied the implementation will attempt to create the protobuf deserializer using the schema registry instead.

Defaulting to Protobuf

If preferred the message type could be set to default as follows:

--message.format=PROTOBUF

Running with Docker

Images are hosted at hub.docker.com/r/obsidiandynamics/kafdrop.

Launch container in background:

docker run -d --rm -p 9000:9000 \
    -e KAFKA_BROKERCONNECT=<host:port,host:port> \
    -e JVM_OPTS="-Xms32M -Xmx64M" \
    -e SERVER_SERVLET_CONTEXTPATH="/" \
    obsidiandynamics/kafdrop

Launch container in background with protobuff definitions:

docker run -d --rm -v <path_to_protobuff_descriptor_files>:/var/protobuf_desc -p 9000:9000 \
    -e KAFKA_BROKERCONNECT=<host:port,host:port> \
    -e JVM_OPTS="-Xms32M -Xmx64M" \
    -e SERVER_SERVLET_CONTEXTPATH="/" \
    -e CMD_ARGS="--message.format=PROTOBUF --protobufdesc.directory=/var/protobuf_desc" \
    obsidiandynamics/kafdrop

Then access the web UI at http://localhost:9000.

Hey there! We hope you really like Kafdrop! Please take a moment to ⭐the repo or Tweet about it.

Running in Kubernetes (using a Helm Chart)

Clone the repository (if necessary):

git clone https://github.com/obsidiandynamics/kafdrop && cd kafdrop

Apply the chart:

helm upgrade -i kafdrop chart --set image.tag=3.x.x \
    --set kafka.brokerConnect=<host:port,host:port> \
    --set server.servlet.contextPath="/" \
    --set cmdArgs="--message.format=AVRO --schemaregistry.connect=http://localhost:8080" \ #optional
    --set jvm.opts="-Xms32M -Xmx64M"

For all Helm configuration options, have a peek into chart/values.yaml.

Replace 3.x.x with the image tag of obsidiandynamics/kafdrop. Services will be bound on port 9000 by default (node port 30900).

Note: The context path must begin with a slash.

Proxy to the Kubernetes cluster:

kubectl proxy

Navigate to http://localhost:8001/api/v1/namespaces/default/services/http:kafdrop:9000/proxy.

Protobuf support via helm chart:

To install with protobuf support, a "facility" option is provided for the deployment, to mount the descriptor files folder, as well as passing the required CMD arguments, via option mountProtoDesc. Example:

helm upgrade -i kafdrop chart --set image.tag=3.x.x \
    --set kafka.brokerConnect=<host:port,host:port> \
    --set server.servlet.contextPath="/" \
    --set mountProtoDesc.enabled=true \
    --set mountProtoDesc.hostPath="<path/to/desc/folder>" \
    --set jvm.opts="-Xms32M -Xmx64M"

Building

After cloning the repository, building is just a matter of running a standard Maven build:

$ mvn clean package

The following command will generate a Docker image:

mvn assembly:single docker:build

Docker Compose

There is a docker-compose.yaml file that bundles a Kafka/ZooKeeper instance with Kafdrop:

cd docker-compose/kafka-kafdrop
docker-compose up

APIs

JSON endpoints

Starting with version 2.0.0, Kafdrop offers a set of Kafka APIs that mirror the existing HTML views. Any existing endpoint can be returned as JSON by simply setting the Accept: application/json header. Some endpoints are JSON only:

• /topic: Returns a list of all topics.

Swagger

To help document the Kafka APIs, Swagger has been included. The Swagger output is available by default at the following Kafdrop URL:

/v2/api-docs

This can be overridden with the following configuration:

springfox.documentation.swagger.v2.path=/new/swagger/path

Currently only the JSON endpoints are included in the Swagger output; the HTML views and Spring Boot debug endpoints are excluded.

You can disable Swagger output with the following configuration:

swagger.enabled=false

CORS Headers

Starting in version 2.0.0, Kafdrop sets CORS headers for all endpoints. You can control the CORS header values with the following configurations:

cors.allowOrigins (default is *)
cors.allowMethods (default is GET,POST,PUT,DELETE)
cors.maxAge (default is 3600)
cors.allowCredentials (default is true)
cors.allowHeaders (default is Origin,Accept,X-Requested-With,Content-Type,Access-Control-Request-Method,Access-Control-Request-Headers,Authorization)

You can also disable CORS entirely with the following configuration:

cors.enabled=false

Topic Configuration

By default, you could delete a topic. If you don't want this feature, you could disable it with:

--topic.deleteEnabled=false

By default, you could create a topic. If you don't want this feature, you could disable it with:

--topic.createEnabled=false

Actuator

Health and info endpoints are available at the following path: /actuator

This can be overridden with the following configuration:

management.endpoints.web.base-path=<path>

Guides

Connecting to a Secure Broker

Kafdrop supports TLS (SSL) and SASL connections for encryption and authentication. This can be configured by providing a combination of the following files (placed into the Kafka root directory):

• kafka.truststore.jks: specifying the certificate for authenticating brokers, if TLS is enabled.
• kafka.keystore.jks: specifying the private key to authenticate the client to the broker, if mutual TLS authentication is required.
• kafka.properties: specifying the necessary configuration, including key/truststore passwords, cipher suites, enabled TLS protocol versions, username/password pairs, etc. When supplying the truststore and/or keystore files, the ssl.truststore.location and ssl.keystore.location properties will be assigned automatically.

Using Docker

The three files above can be supplied to a Docker instance in base-64-encoded form via environment variables:

docker run -d --rm -p 9000:9000 \
    -e KAFKA_BROKERCONNECT=<host:port,host:port> \
    -e KAFKA_PROPERTIES="$(cat kafka.properties | base64)" \
    -e KAFKA_TRUSTSTORE="$(cat kafka.truststore.jks | base64)" \   # optional
    -e KAFKA_KEYSTORE="$(cat kafka.keystore.jks | base64)" \       # optional
    obsidiandynamics/kafdrop

Environment Variables

Basic configuration

Name	Description
KAFKA_BROKERCONNECT	Bootstrap list of Kafka host/port pairs. Defaults to localhost:9092.
KAFKA_PROPERTIES	Additional properties to configure the broker connection (base-64 encoded).
KAFKA_TRUSTSTORE	Certificate for broker authentication (base-64 encoded). Required for TLS/SSL.
KAFKA_KEYSTORE	Private key for mutual TLS authentication (base-64 encoded).
SERVER_SERVLET_CONTEXTPATH	The context path to serve requests on (must end with a /). Defaults to /.
SERVER_PORT	The web server port to listen on. Defaults to 9000.
SCHEMAREGISTRY_CONNECT	The endpoint of Schema Registry for Avro or Protobuf message
SCHEMAREGISTRY_AUTH	Optional basic auth credentials in the form username:password.
CMD_ARGS	Command line arguments to Kafdrop, e.g. --message.format or --protobufdesc.directory or --server.port.

Advanced configuration

Name	Description
JVM_OPTS	JVM options.
JMX_PORT	Port to use for JMX. No default; if unspecified, JMX will not be exposed.
HOST	The hostname to report for the RMI registry (used for JMX). Defaults to localhost.
KAFKA_PROPERTIES_FILE	Internal location where the Kafka properties file will be written to (if KAFKA_PROPERTIES is set). Defaults to kafka.properties.
KAFKA_TRUSTSTORE_FILE	Internal location where the truststore file will be written to (if KAFKA_TRUSTSTORE is set). Defaults to kafka.truststore.jks.
KAFKA_KEYSTORE_FILE	Internal location where the keystore file will be written to (if KAFKA_KEYSTORE is set). Defaults to kafka.keystore.jks.

Using Helm

Like in the Docker example, supply the files in base-64 form:

helm upgrade -i kafdrop chart --set image.tag=3.x.x \
    --set kafka.brokerConnect=<host:port,host:port> \
    --set kafka.properties="$(cat kafka.properties | base64)" \
    --set kafka.truststore="$(cat kafka.truststore.jks | base64)" \
    --set kafka.keystore="$(cat kafka.keystore.jks | base64)"

Updating the Bootstrap theme

Edit the .scss files in the theme directory, then run theme/install.sh. This will overwrite src/main/resources/static/css/bootstrap.min.css. Then build as usual. (Requires npm.)

Securing the Kafdrop UI

Kafdrop UI is an internal service and should be run only on a private & isolated network to eliminate public access. One more layer of security is an authentication mechanism to restrict user access. The authentication mechanism used by Kafdrop is based on Spring Security (default login page: Login Screenshot) with RBAC on top of a database. Here is an example of how to add users. The password is encoded with bcrypt. You can use both PostgreSQL (default) && MySQL (or any other JDBC) for the authentication, in case you want to use the MySQL, mark postgres && unmark mysql both in Maven POM && Application Configuration.

Hey there! We hope you really like Kafdrop! Please take a moment to ⭐the repo or Tweet about it.

Contributing Guidelines

All contributions are more than welcomed. Contributions may close an issue, fix a bug (reported or not reported), add new design blocks, improve the existing code, add new feature, and so on. In the interest of fostering an open and welcoming environment, we as contributors and maintainers pledge to making participation in our project and our community a harassment-free experience for everyone.