Comparisons anchored on Apache Kafka, the Apache-2.0 distributed event streaming platform, KRaft-only since 4.0: against RabbitMQ, Apache Pulsar and Redpanda.
Now that Apache Kafka 4 runs only in KRaft mode, the remaining layout decision is whether servers act as both broker and controller (combined mode) or whether a separate controller quorum runs on dedicated nodes. Covers what the Kafka docs recommend for critical environments, controller quorum sizing, static versus dynamic quorums and when combined mode is acceptable.
Whether to run Apache Kafka on Kubernetes with the Strimzi operators or on virtual machines managed with your own tooling. Covers what Strimzi's Cluster, Topic and User Operators automate, KRaft node pools and dedicated controllers, rolling upgrades, and what a VM deployment leaves to the operations team.
A partitioned event log built for high-volume streaming against a lightweight messaging server with JetStream persistence. Covers delivery guarantees and replay, multi-tenancy through NATS accounts, cluster sizing and replication, connectors (Kafka Connect against NATS bridges), footprint at the edge, and the 2025 agreement that kept NATS under Apache-2.0 in the CNCF.
A distributed log against a message broker: retention and replay versus routing and queues, protocols (Kafka protocol versus AMQP 0-9-1, AMQP 1.0, MQTT, STOMP), replication (KRaft versus quorum queues and streams) and support models.
Two Apache-2.0 streaming platforms compared: Kafka's broker-stored partitions against Pulsar's stateless brokers on BookKeeper, multi-tenancy, geo-replication, release and LTS policies, and operating complexity.
Apache Kafka against a Kafka API-compatible engine without ZooKeeper or the JVM: license (Apache-2.0 versus BSL with enterprise features under a paid key), which features need Redpanda Enterprise, and compatibility for existing clients.
The messaging shortlist compared on messaging model, protocols, replication, multi-tenancy, licenses and support lifecycles, for teams choosing one platform for events and queues.
Log-based event streaming against a JMS and multi-protocol broker: Kafka's partitions, retention and Connect ecosystem against Artemis's queues, JMS and Jakarta Messaging, AMQP and MQTT support, compared for integration with Java enterprise systems and replay needs.
When a Redis stream is enough and when you need Kafka: append-only logs with consumer groups in both, compared on durability and replication, retention, ecosystem, operating effort and Redis's licensing choices since version 8.
Licensing compared for teams weighing Redpanda as a Kafka replacement: Apache 2.0 Kafka against Redpanda's Business Source License core and Redpanda Community License enterprise features, which features need a license key and what happens when it expires.
The three self-hosted messaging systems most compared for microservices: Kafka's log, NATS with JetStream and RabbitMQ's queues and streams, covering delivery guarantees, replication, protocols, footprint and the operating effort of each.
Tell us what you're working on, whether it's a deployment, an audit, a security test or a cyber range. You'll speak with an engineer who can help you scope it.