Confluent REST Proxy for Apache Kafka on Confluent Platform

The Confluent REST Proxy provides a RESTful interface to an Apache Kafka® cluster, making it easy to produce and consume messages, view the state of the cluster, and perform administrative actions without using the native Kafka protocol or clients.

Ready to build?

Sign up for Confluent Cloud with your cloud marketplace account and unlock $1000 in free credits: AWS Marketplace, Google Cloud Marketplace, or Microsoft Azure Marketplace.

Some example use cases are:

  • Reporting data to Kafka from any frontend application built in any language not supported by official Confluent clients

  • Ingesting messages into a stream processing framework that doesn’t yet support Kafka

  • Scripting administrative actions

Security plugin

The REST Proxy Security Plugin is a plugin for REST Proxy that helps authenticate incoming requests and propagates the authenticated principal to requests to Kafka. This plugin enables REST Proxy clients to use the multi-tenant security features of the Kafka broker. Always install this plugin when using REST Proxy in Confluent Enterprise in a production environment. For more information, see Deploy Secure Standalone REST Proxy in Confluent Platform and REST Proxy Security Plugins in Confluent Platform.

REST Proxy Admin REST APIs

The Admin REST APIs allow you to create and manage topics, manage MDS, and produce and consume to topics. The Admin REST APIs are available in these forms:

  • Deploy |cs|, which exposes Admin REST APIs directly on the brokers by default. Confluent Platform ships with Confluent Server.

  • Admin REST APIs are being incrementally added to /cloud/current/, as documented at /cloud/current/.

  • Deploy standalone |crest| nodes, which, in addition to Produce and Consume REST APIs, also offer Admin REST APIs as of API v3.

API availability

There is a difference between the REST APIs available on Kafka brokers deployed with Confluent Server and the REST APIs available with Standalone REST Proxy. Confluent Server provides several REST APIs that are not available in the open-source Apache Kafka® distribution provided with Standalone Kafka REST Proxy. The following REST API endpoints are only available with a Confluent Server deployment:

Configs
/clusters/{cluster_id}/topics/{topic_name}/default-configs
Broker
/clusters/{cluster_id}/brokers:delete

/clusters/{cluster_id}/brokers/{broker_id}:unregister

The GET version of the following endpoint is available in open-source Kafka REST APis. The DELETE option is available only with REST that runs with a Confluent Server deployment.

/clusters/{cluster_id}/brokers/{broker_id}
Cluster Linking
/clusters/{cluster_id}/links
/clusters/{cluster_id}/links/{link_name}
/clusters/{cluster_id}/links/{link_name}/configs
/clusters/{cluster_id}/links/{link_name}/configs/{config_name}
/clusters/{cluster_id}/links/{link_name}/configs:alter
/clusters/{cluster_id}/links/{link_name}/mirrors
/clusters/{cluster_id}/links/-/mirrors
/clusters/{cluster_id}/links/{link_name}/mirrors/{mirror_topic_name}
/clusters/{cluster_id}/links/{link_name}/mirrors:promote
/clusters/{cluster_id}/links/{link_name}/mirrors:failover
/clusters/{cluster_id}/links/{link_name}/mirrors:pause
/clusters/{cluster_id}/links/{link_name}/mirrors:resume
BalancerStatus
/clusters/{cluster_id}/balancer
/clusters/{cluster_id}/balancer/any-uneven-load
BrokerTask
/clusters/{cluster_id}/brokers/-/tasks
/clusters/{cluster_id}/brokers/{broker_id}/tasks
/clusters/{cluster_id}/brokers/-/tasks/{task_type}
/clusters/{cluster_id}/brokers/{broker_id}/tasks/{task_type}
BrokerReplicaExclusion
/clusters/{cluster_id}/broker-replica-exclusions
/clusters/{cluster_id}/broker-replica-exclusions/{broker_id}
/clusters/{cluster_id}/broker-replica-exclusions:create
/clusters/{cluster_id}/broker-replica-exclusions:delete
RemoveBrokerTask
/clusters/{cluster_id}/remove-broker-tasks
/clusters/{cluster_id}/remove-broker-tasks/{broker_id}
ReplicaStatus
/clusters/{cluster_id}/topics/-/partitions/-/replica-status
/clusters/{cluster_id}/topics/{topic_name}/partitions/-/replica-status
/clusters/{cluster_id}/topics/{topic_name}/partitions/{partition_id}/replica-status

REST that runs with a Confluent Server deployment provides the full set of REST APIs. REST that runs in a Standalone deployment consists of the open-source Kafka REST APIs only. For more information about the open-source Kafka REST APIs available, see Kafka REST Proxy and the openapi yaml.

Features

The following features are currently exposed and available through Confluent REST.

  • Metadata - Read most metadata about the cluster, including brokers, topics, partitions, and configurations, with GET requests for the corresponding URLs.

  • Producers - Instead of exposing producer objects, the API accepts produce requests targeted at specific topics or partitions and routes them all through a small pool of producers.

    • Producer configuration - Producer instances share resources, so you cannot set configuration files on a per-request basis. Adjust settings globally by passing new producer settings in the REST Proxy configuration. For example, you might pass in the compression.type option to enable site-wide compression to reduce storage and network overhead.

  • Consumers - Consumers are stateful and tied to specific REST Proxy instances. Offset commit can be automatic or explicitly requested by the user. Currently limited to one thread per consumer; use two or more consumers for higher throughput. The REST Proxy uses either the high level consumer, API v1, or the new 0.9 consumer, API v2, to support consumer-groups that can read from topics. Note: use v2 API instead of the v1 API.

    • Consumer configuration - Consumer instances do not share resources, though they share underlying server resources. Limited configuration options are available through the API. Adjust settings globally by passing consumer settings in the REST Proxy configuration.

  • Data formats - The REST Proxy can read and write data using JSON, raw bytes encoded with base64, or JSON-encoded Avro, Protobuf, or JSON Schema. With Avro, Protobuf, or JSON Schema, the system registers schemas and validates them against Schema Registry.

  • |crest| clusters and load balancing - The REST Proxy supports two or more instances running together to spread load and can safely run behind load balancing mechanisms, for example round robin DNS, discovery services, or load balancers, as long as instances are configured correctly.

  • Simple consumer - Use the high-level consumer. Use low-level read operations only for specific cases, such as retrieving messages at specific offsets.

See also

For an example that shows how to set Docker environment variables for Confluent Platform running on KRaft, see the Confluent Platform demo. Refer to the demo’s docker-compose.yml file for a configuration reference.

Here’s a list of features that aren’t yet supported:

  • Multi-topic Produce Requests - Currently each produce request can address only a single topic or topic-partition. Most use cases do not require multi-topic produce requests; they add complexity to the API, and clients can easily split data across requests if necessary

  • Most Producer/Consumer Overrides in Requests - Only a few key overrides are available in the API. Administrators can set global overrides. The reason is twofold. First, proxies are multi-tenant, so most user-requested overrides need restrictions to ensure they do not impact other users. Second, tying the API too much to the implementation restricts future API improvements; this is important with the new upcoming consumer implementation.

  • Allow different serializer for key and value - Currently, REST Proxy chooses the serializer based on the Content-Type header; as a result, the serializer for key and value must be the same in this design.

Installation

Before starting the REST Proxy, start Kafka and Schema Registry. For installation instructions, see Install Confluent Platform On-Premises.

Deployment

Starting the Confluent REST Proxy service is simple after its dependencies are running:

# Start the REST Proxy. The default settings automatically work with the
# default settings for local Kafka nodes.
<path-to-confluent>/bin/kafka-rest-start etc/kafka-rest/kafka-rest.properties

If you installed Debian or RPM packages, run kafka-rest-start as it will be on your PATH. The kafka-rest.properties file contains configuration settings. The default configuration included with the REST Proxy has convenient defaults for a local testing setup and should be modified for a production deployment. By default, the server starts bound to port 8082 and does not specify a unique instance ID (required to safely run multiple proxies concurrently).

If you started the service in the background, use the following command to stop it:

bin/kafka-rest-stop

Development

To build a development version, you might need development versions of common, rest-utils, and schema-registry. After installing these, build the Confluent REST Proxy with Maven. All the standard lifecycle phases work. During development, use

mvn -f kafka-rest/pom.xml compile

to build,

mvn -f kafka-rest/pom.xml test

to run the unit and integration tests, and

mvn exec:java

to run an instance of the proxy against a local Kafka cluster (using the default configuration included with Kafka).

To create a packaged version, optionally skipping the tests:

mvn -f kafka-rest/pom.xml package [-DskipTests]

This produces a version ready for production in target/kafka-rest-$VERSION-package containing a directory layout similar to the packaged binary versions. Produce a standalone fat jar using the standalone profile:

mvn -f kafka-rest/pom.xml package -P standalone [-DskipTests]

generating target/kafka-rest-$VERSION-standalone.jar, which includes all the dependencies as well.

To run a local Kafka and REST Proxy cluster, for testing:

./testing/environments/minimal/run.sh

Suggested Resources