Issues

Dynamic multitenancy configuration

Update ORM mapper tests relying on entity graphs

See if `java.util.stream.Collector` can be used instead of `ProjectionCollector`

Read-only indexes

Jakarta Data implementation / annotation processor

Expose flight recorder events

Benchmarks for async indexing

Add an alternative mass loading strategy that does not rely on identifier loading

Hibernate ORM's embeddable inheritance with Hibernate Search ORM mapper

ORM null throwing out current batch from massindex (due to ORM loader making entity==null)

SPI to register/use a separate PojoIndexingProcessor to index the identifier

Automated testing for Amazon OpenSearch Serverless

`fromRoot` projection

Allow passing bean references to binders/bridges in the programmatic API

Detect ValueBridge applied to composite objects and warn or fix the dependency model

Don't require bi-directional associations for @IndexedEmbedded/bridges that extract the ID from the target of an association on the owning side

Take into account immutable properties/types when resolving dependencies for reindexing resolvers

Replace remaining usage of URL constructors with URI when switching to JDK20+ as the baseline

Factorize all code related to POJO tree processing

Do not reindex when irrelevant changes occur inside embedded objects

Inner hits in search queries

Annotated-defined projection with values injected through setters instead of constructor

Validation/caching of @ProjectionConstructor on startup

SPI for "hot" restart that allows compatible schema changes with Lucene but doesn't interrupt search/indexing availability

Log a warning or throw an error for Hibernate Search annotations in the wrong place

Primary/replica setup for the Lucene backend with outbox-polling coordination

Extraction of data from entities on session flush leads to out-of-sync index if entities were just recently merged, because @OneToMany associations are still null

Filter data by tenant when providing a low-level LuceneIndexReader

Set and validate the boolean query max clause count

Object fields that are both nested and flattened (ES's include_in_parent)

Custom names for metadata fields with Elasticsearch

Custom container extractors with the ORM mapper

Provide integrations to real-world spatial libraries

Restore support for configuring a custom LockFactory

Implement multi-index operations such as purge or delete-by-query as a single Elasticsearch call

Allow indexing/query works to be cancelled upon timeout or by the user

Improve the mass indexing monitor count in case of errors

Predicates with reduced scope (narrowed target type) in predicate DSL

Add smoke tests for each module and each type of bundle

Explicit checks when referencing Elasticsearch analysis definitions

Declared parameters for Search queries

Projections on bridge-registered object fields

Dynamic index "partitioning"

POJO mapper annotations to configure backend-specific, index-scoped options

Add a bytecode enhancement feature checking properties accessed from within a bridge were declared as dependencies

Improve the accessor-based API in bridges and promote it to non-experimental

Restore support for OSGi

Geoshape search

Support automatic reindexing when an association has multiple inverse sides