kafkactl

Command-line interface for Apache Kafka

TLDR

List topics

$ kafkactl get topics

Describe a topic

$ kafkactl describe topic [my-topic]

Create a topic with partitions

$ kafkactl create topic [my-topic] --partitions [32]

Consume from the beginning of a topic

$ kafkactl consume [my-topic] --from-beginning

Consume the last N messages

$ kafkactl consume [my-topic] --tail=[5]

Produce a message with key and value

$ kafkactl produce [my-topic] --key=[my-key] --value=[my-value]

Produce from stdin with a custom separator

$ echo "[key]#[value]" | kafkactl produce [my-topic] --separator=#

List consumer groups

$ kafkactl get consumer-groups

Describe a consumer group

$ kafkactl describe consumer-group [my-group]

SYNOPSIS

kafkactl [*global-options*] *command* [*args*]

kafkactl is a command-line interface for interacting with Apache Kafka clusters. It covers common day-to-day tasks: listing and managing topics, consuming and producing messages, inspecting consumer groups, and working with schema registry encodings (Avro, JSON Schema, Protobuf).Contexts in the config file let you switch between multiple clusters. Optional Kubernetes integration can run kafkactl inside a cluster so you can reach brokers that are only reachable via kubectl. Shell completion is available for bash, zsh, and fish, including dynamic completion for topics and consumer groups.Default config path is $HOME/.config/kafkactl/config.yml (also looked up under $HOME/.kafkactl/, $APPDATA/kafkactl/, and /etc/kafkactl/). Project-level config files near a git root are supported. Credentials can be prompted interactively and stored in the OS keyring, or supplied via config and environment variables.

PARAMETERS

get topics

List topics on the current context.

describe topic *topic*

Show topic details; --all-configs includes all topic configs; --skip-empty omits empty partitions.

create topic *topic*

Create a topic. Options include --partitions, --replication-factor, --config *key=value*, and --file for cloning from exported JSON.

alter topic *topic*

Change partitions, replication factor, or topic configs.

consume *topic*

Consume messages. Common flags: --from-beginning, --tail=N, --exit, --group, --print-keys, --print-timestamps, --print-headers, --print-partitions, --from-timestamp, --to-timestamp, --max-messages, --filter-key, --filter-value, -o yaml|json.

produce *topic*

Produce messages via --key / --value, stdin, or --file. Supports --separator, --input-format=json, --header, encodings, protobuf types, and --rate.

get consumer-groups

List consumer groups; optional --topic filter.

describe consumer-group *group*

Show group details and lag; --only-with-lag filters.

completion bash|zsh|fish

Generate shell completion scripts.

--config-file=path

Use an alternate config file.

See kafkactl --help and the online command docs for the full command tree and flags.

CONFIGURATION

Create $HOME/.config/kafkactl/config.yml with one or more named contexts (brokers, TLS, SASL, schema registry, optional kubernetes). If no file exists, a default suitable for a local single-node cluster is generated. Every config key can be overridden by environment variables. See upstream docs for the full schema.

INSTALL

brew install kafkactl

nix profile install nixpkgs#kafkactl

CAVEATS

--from-timestamp / --to-timestamp compute offsets at consumer start; they are not wall-clock schedulers for future times. Producing with wrong encoding or schema versions can write data consumers cannot decode. When using Kubernetes mode, kubectl access and RBAC must allow the configured operations.

RESOURCES

Source code · Documentation