In the Kafka world nothing is easy, but esque (pronounced esk) is an attempt at it.

esque is a user-centric command line interface for Kafka administration.

Some stuff is hard, and that is okay, but listing your kafka topics shouldn't be.

While adopting kafka at real.digital(Kaufland) we noticed the immense entry barrier it poses to newcomers. We can't recount how often we wrote Slack messages asking for the script to check the status of topics or consumer groups. This is partly (but not only) due to a fragmented and unclear definition of tooling and APIs for kafka. In a wide array of administration tools, esque distances itself by striving to provide Kafka Ops for Humans, in a usable and natural way.

We feel that the goal of esque embodies the principle: “keep easy things easy, and make hard things possible”.

• batteries included
• feature rich
• robust
• insightful
• by engineers for engineers

• Support for any type of Kafka deployment >1.2
• Display Resources (Topics, Consumer Groups, Brokers)
• Get detailed Overviews of Resources (Topics, Consumer Groups, Brokers)
• Create/Delete Topics
• Edit Topic Configurations
• Edit Consumer Offset for Topics
• SASL/SSL Support out of the box
• Consuming from Avro,Protobuf,plaintext,struct Topics (including Avro Schema Resolution from Schema Registry)
• Producing to and from Avro and Plaintext Topics (including Avro Schema Resolution from Schema Registry)
• Context Switch (Easily Switch between pre-defined Clusters)
• Kafka Ping (Test roundtrip time to your kafka cluster)

$ esque
Usage: esque [OPTIONS] COMMAND [ARGS]...

  esque - an operational kafka tool.

  In the Kafka world nothing is easy, but esque (pronounced esk) is an
  attempt at it.

Options:
  --version      Show the version and exit.
  -v, --verbose  Return stack trace on error.
  --no-verify    Skip all verification dialogs and answer them with yes.
  --help         Show this message and exit.

Commands:
  apply      Apply a set of topic configurations.
  config     Configuration-related options.
  consume    Consume messages from a topic.
  create     Create a new instance of a resource.
  ctx        List contexts and switch between them.
  delete     Delete a resource.
  describe   Get detailed information about a resource.
  edit       Edit a resource.
  get        Get a quick overview of different resources.
  io         Run a message pipeline.
  ping       Test the connection to the kafka cluster.
  produce    Produce messages to a topic.
  set        Set resource attributes.
  urlencode  Url-encode the given value.

esque is available at pypi.org and can be installed with pip install esque.

esque requires Python 3.9+ to run(Python 3.13 is not yet supported)..

There are no wheels for Alpine Linux, so esque requires a few extra dependencies to build them during installation.

apk add python3-dev py3-pip librdkafka librdkafka-dev g++

The installation for Kafka is slightly different for Apple Silicon devices, so simply running pip install esque may result in errors.

The fix for this is to first install the librdkafka library with brew install librdkafka (make note of which version is installed). Then, add the following to the .zshrc file with the correct version of librdkafka:

export C_INCLUDE_PATH=/opt/homebrew/Cellar/librdkafka/X.X.X/include/
export LIBRARY_PATH=/opt/homebrew/Cellar/librdkafka/X.X.X/lib/

After starting a new shell (or source ~/.zshrc), you will be able to install esque normally with pip.

The autocompletion scripts for bash and zsh can be generated by running esque config autocomplete.

When starting esque for the first time the following message will appear:

No config provided in ~/.esque
Should a sample file be created in ~/.esque [y/N]:

When answering with y esque will copy over the sample config to ~/.esque/esque_config.yaml. Afterwards you can modify that file to fit your cluster definitions.

Alternatively might just provide a config file following the sample config's file in that path.

version: 1
current_context: local
contexts:
  # This context corresponds to a local development cluster
  # created by docker-compose when running esque from the host machine.
  local:
    bootstrap_servers:
      - localhost:9092
    security_protocol: PLAINTEXT
    schema_registry: http://localhost:8081
    proto:
      topic1:
        protoc_py_path: /home/user/pb/
        module_name: hi_pb2
        class_name: HelloWorldResponse
      any_key:
        protoc_py_path: 'path to compiled files using protoc'
        module_name: "api_stubs_py.api.bye_pb2"
        class_name: ByeMessage
    default_values:
      num_partitions: 1
      replication_factor: 1

The config for the apply command has to be a yaml file and is given with the option -f or --file.

In the current version only topic configurations can be changed and specified.

It has to use the same schema, which is used for the following example:

topics:
  - name: topic_one
    replication_factor: 3
    num_partitions: 50
    config:
      cleanup.policy: compact
  - name: topic_two
    replication_factor: 3
    num_partitions: 50
    config:
      cleanup.policy: compact

To de-serialize protobuf files you need to have generated files using protoc command. Let's assume you have following proto file and you have dispatched message to a kafka topic with HelloWorldResponse:

syntax = "proto3";

service Hi {
    rpc Get (HelloWorldRequest) returns (HelloWorldResponse);
}

message HelloWorldRequest {
    string name = 1;
}

message HelloWorldResponse {
    string type_string = 1;
    optional string optional_string = 2;
    EnumType type_enum = 3;
    int32 type_int32 = 4;
    int64 type_int64 = 5;
    optional int64 optional_int64 = 6;
    float type_float = 7;
}
enum EnumType {
    ENUM_TYPE_UNSPECIFIED = 0;
}

first you need to compile it :

protoc --python_out /home/user/pb/ hi.proto

then it will create files like :

hi.proto
hi_pb2.py

After compiling protobuf files for python you need three things.

• protoc_py_path: absolute path to the compiled files using protoc. example : /home/user/pb/
• module_name: path to the file having contains your message. our example: hi_pb2 (without .py). usually you can find it the pb_2 file in line starts like _builder.BuildTopDescriptorsAndMessages(DESCRIPTOR, [module_name], globals())
• class_name: message name in protobuf. our example: HelloWorldResponse

You can insert this parameters using consume command. example :

esque consume topic_name -s proto --val-protoc-py-path /home/user/pb/ --val-protoc-module-name hi_pb2 --val-protoc-class-name HelloWorldResponse

if this is a topic you use everyday you can save this in configuration file like :

    proto:
      topic_name:
        protoc_py_path: /home/user/pb/
        module_name: hi_pb2
        class_name: HelloWorldResponse

and next you just need to run esque consume topic_name -s proto.

If you have same configuration for another topic you don't need to copy the same configuration. you just need to use --val-proto-key to specify the key in configuration file. for example if your new topic is called topic2 then your command will be : esque consume topic_name -s proto --val-proto-key topic_name

In this examples you have assumed you want to de-serialize proto messages from value of kafka topic. if you have the same thing but in key then all parameters will be changed from --val- prefix to --key- prefix.

To setup your development environment, make sure you have at least Python 3.9 & poetry installed, then run

poetry install
poetry shell

To install pre commit hooks run:

pip install pre-commit
pre-commit install
pre-commit install-hooks

esque comes with a docker-compose based kafka stack which you can start up with make test-suite.

You can then run the integration tests against this stack with pytest tests/ --integration --local.

Alternatively you can go the fast way and just run the whole stack + integration tests in docker:

make integration-test

If you only want the unit tests, just run:

make test