Hi all, Chukwa developers have decided to throw in the towels, and send Chukwa to Apache Attics. In the past 10 years, original Chukwa design pattern has become one of the most commonly used pattern for running reports on Big Data. Each component can be take apart and re-engineer for better performance. Many frameworks have sprouts to improve every aspect of big data analytics. Kafka, Prometheus and Nifi have gotten better at data collection than Chukwa agent/collectors. Parquet, HBase have gotten better at storing compressed semi structure data than Hadoop RecordIO. Hive, Spark, and Phoenix have gotten better at relational analysis than Mapreduce + Chukwa MDL. Grafana, Tableau are getting better at data visualization than Chukwa HICC. Instead of a monolithic big data analytics framework, Chukwa like system can be built from open source frameworks. Chukwa may have inspired others to do better, or maybe it didn't. We thank you for all your support over the years.
We have concluded that it would be irresponsible to keep our users and developers waiting for a framework that is readily available in open source. We wish you all the best in big data adventures. Farewell... regards, Eric
