Release notes

Fixed an Unknown QueryBuilder error when a GeoPoint query was used inside a join request.

Fixed a cache issue that arose due to the shard life-cycle not being considered during fingerprint computation.

Fixed a NullPointerException when an error occurred and the cause message was null.

Fixed a bug caused by using an unsupported operation of exist in the aggregation context with projected fields.

Fixed a concurrent modification exception in a multi-search request.

Fixed a bug caused by skipping the pruning step when an error occurred during the request planning phase.

Improved concurrency management on multi-search requests.

Upgraded to Elasticsearch 8.14.3.

Fixed a resource leak that occurred during job initialization.

Fixed a race condition in the data staging service.

Added a new setting, federate.internal.actions.timeout, to configure a timeout for internal actions, with a default value of 30s.

Introduced a new node attribute federate.enable which allows you to exclude nodes from search and computation when it is set to false.

Enhanced Elasticsearch aggregations to work with Federate joins in static and adaptive planner requests.

Added multi-search request capabilities for Adaptive Query Planner (AQP).

Upgraded to Elasticsearch 8.14.1.

Improved function_score to accept join in the filters scope.

Added a push down optimization to queries inside aggregations for HASH_JOIN.

Added the following values to the feature flag parameter (siren-federate-features) to manipulate the push down optimization in aggregations:

Improved the documentation on thread pools and its types.

Improved the documentation on settings.

Increased the default value of federate.planner.multi_conditional.index_lookup.limit to 50,000.

Improved the pruning in AQP to avoid duplicated lookup requests.

Fixed join query failures caused by unsupported matches API used by highlighters.

Fixed an error message when using a plain string instead of an array for the on clause.

Fixed an issue with the caching of AQP requests caused by the wrong searchLock selection.

Fixed the explain API to not prune tasks anymore.

Fixed a bug that prevented the detection of EmptyJoin inside aggregations.

Fixed a bug that prevented federate_object from referencing fields projected from a child scope.

Fixed a bug in the REST _siren/cache/planner endpoint to handle also the field_caps cache used during the planner phase.

Marked REST endpoint _siren/cache as deprecated, use _siren/cache/planner instead.

Marked connector APIs as deprecated; connector APIs includes:

Deprecated the setting federate.planner.index_join.max_lookups. The new setting to use is federate.planner.multi_conditional.index_lookup.limit.