This 62nd edition of the Kafka Monthly Digest covers what happened in the Apache Kafka community in March 2023.
For last month’s digest, see Kafka Monthly Digest: February 2023.
Releases
There is one release in progress: 3.5.0.
3.5.0
The release process for 3.5.0 continued. KIP freeze happened on March 22. The next milestone is feature freeze on April 12. You can find the release plan with all the details in the wiki.
KIPs
Last month, the community submitted 8 KIPs (KIP-910 to KIP-917). I'll highlight a few of them:
- KIP-911: Add source tag to MirrorSourceConnector metrics. Currently the metrics from MirrorMaker Source connector don't include the source cluster alias as a label, they have the topic, partition and target cluster alias. This KIP introduces a new configuration,
add.source.alias.to.metrics
to add the source cluster alias as a label. From Kafka 4.0, the source cluster alias will always be added. - KIP-914: Join Processor Semantics for Versioned Stores. Recently KIP-889 introduced versioned state stores in Kafka Streams. This KIP proposes updating the semantics of stream-table and table-table joins to leverage versioned stores and provide more accurate results.
- KIP-915: Txn and Group Coordinator Downgrade Foundation. Brokers serve as coordinators for consumer groups and transactional producers. When there's a major upgrade to coordinators like with KIP-848, it can be hard to downgrade back to a previous version. This KIP aims at bringing mechanisms to enable downgrades. There's a follow up KIP, KIP-917, that builds on top of this to make the group coordinator fully backward and forward compatible.
Community Releases
- strimzi-kafka-operator 0.34: Strimzi is a Kubernetes Operator for running Kafka. It adds support for Kafka 3.4.0 and now requires at least Kubernetes 1.19. This new version assigns stable identities to Connect pods to avoid rebalances during common maintenance operations like rolling restarts.
- franz-go 1.13: Franz-go is a pure Go library for Kafka. It now requires Go 1.18. It improves how non retryable errors are handled and includes a number of other small improvements and bug fixes.
Blogs
I selected some interesting blog articles that were published last month:
- Testing Kafka-based Asynchronous Workflows Using OpenTelemetry
- Does Apache Kafka® really preserve message ordering?
To learn more about Kafka, visit Red Hat Developer's Apache Kafka topic page.