The Apache Druid team is proud to announce the release of Apache Druid 35.0.0.
Druid is a high performance analytics data store for event-driven data.
Apache Druid 35.0.0 contains over 220 new features, performance
enhancements, bug fixes, and major documentation improvements from 29
contributors. Major new features and improvements include:
- upgraded Jetty to 12 (from 9)
- official support for Java 21, and dropped support for Java 11
- multi-stage query engine is now a core feature instead of an extension
- new 'lines' input format which can read newline delimited text for
later processing (such as TRY_PARSE_JSON)
- improvements to experimental projections feature to support filtered
projections, along with several bug fixes
- experimental 'virtual storage fabric' mode for historical servers
- new 'druid-exact-count-bitmap' contrib extension to perform count
distinct over a long column using roaring bitmaps
- several query performance enhancements and much more
Source and binary distributions can be downloaded from:
https://druid.apache.org/downloads.html
Release notes are at:
https://github.com/apache/druid/releases/tag/druid-35.0.0
A big thank you to all the contributors in this release!