Kafta is a modern non-JVM command line for managing Kafka clusters

After several nights, weekends and late nights. Today we have the first stable release of Kafta. This project was born a while ago with me and @snakeice. We spent several days frustrated using the native commands that come with Kafka.

Kafta was created by developers for developers. We feel the pain of maintaining a kafka cluster using the bash's provided by apache-kafka, it's confusing and the experience is miserable. To facilitate the adoption of the kafka, the kafta began to be born. Kafta is a golang project that is easy to install, easy to configure and simple to use.

Overview

Kafta is built on a structure of commands, arguments & flags. Kafta will always be interacting on one cluster at a time, the reason for this is not having to pass which cluster is in each command, as it is with most command lines for kafka.

To see all exists commands, run:

$ kafta 
Usage:
  kafta [command]

Available Commands:
  broker      broker management
  cluster     cluster management
  completion  Output shell completion code 
  config      Modify config files
  console     Console management
  consumer    Consumer group management
  help        Help about any command
  schema      Schema Registry management
  topic       Topics management
  version     Print the kafta version

Flags:
  --context string       The name of the kafkaconfig context to use
  -d, --debug                Debug mode
  -h, --help                 help for kafta
      --kafkaconfig string   Path to the kafkaconfig file to 

Use "kafta [command] --help" for more information about a command.

To see all options exists relate to same command, run:

$ kafta topic
Topics management

Usage:
  kafta topic [command]

Available Commands:
  create      Create topics
  delete      Delete topics
  describe    Describe a topic
  list        List topics

Simple commands

Kafta is very similar to other cli's, it was made to avoid thinking "what is the syntax for this command". A great example is creating a topic. It's so simple, just run this command

$ kafta topic create my-topic --rf 3 --partitions 10
Topic created

That's it! your topic is created.

There are default values ​​for partition and replication factor, which is why it can only be used without specifying RF or partition. The topic will be created with RF=3 and partition=10. Example:

$ kafta topic create my-topic
Topic created

Installing

Use go get to install the latest version. This command will install the Kafta executable along with the library and its dependencies:

go < 1.18: go get -u github.com/electric-saw/kafta
go >= 1.18: go install github.com/electric-saw/kafta/cmd/kafta@latest

If you prefer, just download the binary and run it on your machine, however and wherever you want.

Contexts & Config

Kafta will create a config file in ~/.kafta/config. This yaml is used to support kafka multi-clusters and avoid passing all addresses every time.

Each cluster in Kafta is called a context, Kafta's proposal is to be more than a simple Kafka manager, thinking about managing schema-registry, connect and other parts of a Kafka environment, we call this group a context.

To set up a new context, create a new config via Kafta, you'll need to provide some information, don't worry, it's all in terminal, you don't need to edit any XML \o/

Follow the example:

$ kafta config set-context production
Bootstrap servers: b-1.mydomain:9092,b-2.mydomain:9092,b-3.mydomain:9092
Schema registry: https://schema-registry.com
Use SASL: y
SASL Algorithm: sha512
User: myuser
✔ Password: ******

To list the contexts, run:

$ kafta config get-contexts
+---------------+---------------------------+-----------------------------+------+---------+
| NAME          | CLUSTER                   | SCHEMA REGISTRY             | KSQL | CURRENT |
+---------------+---------------------------+-----------------------------+------+---------+
| dev           | b-1.mydomain:9092         | https://schema-registry.com |      | true    |
| production    | b-3.productiondomain:9092 | https://schema-registry.com |      | false   |
+---------------+---------------------------+-----------------------------+------+---------+

This part is a differential of Kafta. It is designed for environments with many clusters and it is essential to be easy to move from one cluster to another. To change the current cluster, run:

$ kafta config use-context production
Switched to context "production".

Others commands

Kafta has several commands and you can find an example of them in the project's README. Some commands it has:

• Consumer Groups
• Schema Registry - Partial support
• Consumer/Producer - Thanks Vinicius Folgosa for this contribution
• Broker
• Cluster configs
• Topics

Contribution

Kafta is very new and we have many opportunities to change and create many things, if you are interested in making a feature, please open a new issue and start a conversation with us. If you don't have time for that, just share the project and click on the star if you liked it ;)

Comments

[Christophe Colombier]

I used kcat and topicctl

github.com/edenhill/kcat
github.com/segmentio/topicctl

What do think about it if you know them ?

[Lucas]

kcat I had several problems configuring it in Arch/fedora. Topicctl is good but it is very topic focused and Kafta for now has a big focus on management