Release notes



  • Upgraded Apache Arrow to 11.0.0.

  • Improved the error message sent to the client when trying to create a point in time (PIT) in virtual indices.

Bug fixes

  • Fixed the preservation of the thread context for queued data uploads.

  • Fixed an issue of wrongly cached PIT queries.



  • Upgraded to Elasticsearch 8.7.1

Bug fixes

  • Fixed an issue where there was a need to explicitly specify the INDEX_JOIN strategy when performing a multi-conditional join on a single node cluster.

  • Fixed the calculation of debug information for the throughput value, which was previously based on the output data size but is now calculated using the input data size.



  • Added multi-conditional join capabilities.

  • Added new binary operators to be used in a join (gt, lt, gte and lte).


  • Upgraded to Elasticsearch 8.7.0

  • Adapted ROUTING_JOIN to work on Adaptive Query Planner (AQP).

  • Adapted runtime fields to work on AQP.

  • Adapted Point in Time (PIT) to work on AQP.

  • Enhanced performance by removing several blocking Java calls.

  • Improved documentation on Query DSL and PIT limitations.

Bug fixes

  • Resolved a race condition involving job cleanup requests and data packet release.

  • Fixed an integer overflow issue with batch size computation.

  • Fixed ArrowBuf initialization to be in a privileged block.

Breaking changes

  • Plugin initialization will fail if memory is incorrectly configured.

    • Java Max Direct Memory (XX:MaxDirectMemorySize) must be greater than Siren memory root limit (siren.memory.root.limit).

Known issues

  • When performing a multi-conditional join on a single node cluster, it is necessary to explicitly specify the INDEX_JOIN strategy.