New in Confluent Cloud: Making Data & Pipelines Accessible for AI-Ready Streaming | Learn More

Login Contact Sales

Watch DemoTry Confluent
Watch DemoTry Confluent

Log Compaction | Highlights in the Kafka and Stream Processing Community | November 2015

Written By

kafka-logo-wide (1)

The Apache Kafka community just concluded its busiest month ever. As we are preparing for the upcoming release of Kafka 0.9.0.0, the community worked together to close a record number of bugs and added quite a few last-minute improvements to new features. Here’s a summary of some of the community highlights:

  • SASL/Kerberos-based authentication was added to Kafka and will be included in the 0.9.0.0 release
  • The new Kafka Consumer has a client-side partition-assignment support, allowing consumer groups to implement their own partition assignment strategy, in addition to those provided in Apache Kafka. You can read the very detailed design document or take a look at the patch.
  • Huge improvements on the Copycat export/import feature – Copycat now works in distributed mode and there is a nice REST API too.
  • The community also worked hard on making Kafka more testable. In the last few month the old spagetti-like system tests slowly got replaced by the more modular Ducktape-based integration tests. Last week we removed the old tests completely and moving forward we will only maintain the new test suite.
  • This is the month of the Kafka Meetup! New Kafka meetups are popping up in cities all over the world including Seattle, Austin and London. I encourage you to attend if you happen to be in one of those cities.
  • If you are a Kafka internals geek, you’ll enjoy reading the deep details behind Kafka’s delayed response mechanism.
  • For those interested in the latest in stream processing, Data Artisans shared interesting highlights from the first Apache Flink conference.
  • Remember that the Kafka Summit “Call for Proposals” is open so make sure you get your entry in before the January 11, 2016 deadline. If you’re planning to attend and have ideas of sessions you’d like to see on the program tweet us @ConfluentInc #kafkasummit to let us know.

If you are interested in contributing to Apache Kafka, check out our contributor guide to help you get started.

Apache Kafka

Download the Confluent Platform 

  • Gwen Shapira is a Software Enginner at Confluent. She has 15 years of experience working with code and customers to build scalable data architectures, integrating relational and big data technologies. She currently specialises in building real-time reliable data processing pipelines using Apache Kafka. Gwen is an Oracle Ace Director, an author of books including “Kafka, the Definitive Guide”, and a frequent presenter at data related conferences. Gwen is also a committer on the Apache Kafka and Apache Sqoop projects.

Did you like this blog post? Share it now

Proven: Up to 73% Fewer Brokers With Confluent Private Cloud – and More

Proven: Up to 73% Fewer Brokers With Confluent Private Cloud – and More

Confluent Private Cloud matches Kafka performance with up to 73% fewer brokers. Benchmarks prove massive TCO savings, while Centralized Policy Enforcement streamlines data governance and broker-native multi-tenancy provides the isolation needed to end "noisy neighbor" issues for good.

More Signal, Less Guesswork: New Kafka Observability Updates in Confluent Cloud

More Signal, Less Guesswork: New Kafka Observability Updates in Confluent Cloud

Confluent Cloud rolled out new observability updates that give operators direct visibility into streaming workload performance. New Metrics API signals expose client throttling by principal, consumer group rebalance duration, connection attempt spikes, and compacted partition counts.