Forward port release notes for v8.15.3 (#115029)

Author: elasticsearchmachine

docs/reference/release-notes.asciidoc

@@ -7,6 +7,7 @@
 This section summarizes the changes in each release.
 
 * <<release-notes-8.16.0>>
+* <<release-notes-8.15.3>>
 * <<release-notes-8.15.2>>
 * <<release-notes-8.15.1>>
 * <<release-notes-8.15.0>>
@@ -74,6 +75,7 @@ This section summarizes the changes in each release.
 --
 
 include::release-notes/8.16.0.asciidoc[]
+include::release-notes/8.15.3.asciidoc[]
 include::release-notes/8.15.2.asciidoc[]
 include::release-notes/8.15.1.asciidoc[]
 include::release-notes/8.15.0.asciidoc[]

docs/reference/release-notes/8.15.3.asciidoc

@@ -0,0 +1,47 @@
+[[release-notes-8.15.3]]
+== {es} version 8.15.3
+
+Also see <<breaking-changes-8.15,Breaking changes in 8.15>>.
+
+[[bug-8.15.3]]
+[float]
+=== Bug fixes
+
+Aggregations::
+* Don't validate internal stats if they are empty {es-pull}113846[#113846] (issue: {es-issue}113811[#113811])
+* Fix `needsScore` computation in `GlobalOrdCardinalityAggregator` {es-pull}113129[#113129] (issue: {es-issue}112975[#112975])
+
+Authentication::
+* Enables cluster state role mapper, to include ECK operator-defined role mappings in role resolution {es-pull}114337[#114337]
+
+ES|QL::
+* ES|QL: Ensure minimum capacity for `PlanStreamInput` caches {es-pull}114116[#114116]
+* ES|QL: Skip CASE function from `InferIsNotNull` rule checks {es-pull}113123[#113123] (issue: {es-issue}112704[#112704])
+* [ESQL] Fix init value in max float aggregation {es-pull}113699[#113699]
+* [ESQL] Support datetime data type in Least and Greatest functions {es-pull}113961[#113961]
+
+Machine Learning::
+* Fix check on E5 model platform compatibility {es-pull}113437[#113437] (issue: {es-issue}113577[#113577])
+* Handle parsing ingest processors where definition is not a object {es-pull}113697[#113697] (issue: {es-issue}113615[#113615])
+* [ML][backport] Warn for model load failures if they have a status code <500 {es-pull}113410[#113410]
+* [M] Fix error message formatting {es-pull}113266[#113266]
+
+Search::
+* Fix analyzed wildcard query in simple_query_string when disjunctions is empty {es-pull}114264[#114264] (issue: {es-issue}114185[#114185])
+* Fix collapse interaction with stored fields {es-pull}112761[#112761] (issue: {es-issue}112646[#112646])
+
+[[enhancement-8.15.3]]
+[float]
+=== Enhancements
+
+Machine Learning::
+* Write downloaded model parts async {es-pull}111684[#111684]
+
+[[upgrade-8.15.3]]
+[float]
+=== Upgrades
+
+Snapshot/Restore::
+* Upgrade protobufer to 3.25.5 {es-pull}113869[#113869]
+
+

docs/reference/release-notes/highlights.asciidoc

@@ -102,6 +102,31 @@ index codec best compression. (#112665)
 
 {es-pull}112857[#112857]
 
+[discrete]
+[[esql_introduce_per_agg_filter]]
+=== ESQL: Introduce per agg filter
+Add support for aggregation scoped filters that work dynamically on the
+data in each group.
+
+[source,esql]
+----
+| STATS success = COUNT(*) WHERE 200 <= code AND code < 300,
+        redirect = COUNT(*) WHERE 300 <= code AND code < 400,
+        client_err = COUNT(*) WHERE 400 <= code AND code < 500,
+        server_err = COUNT(*) WHERE 500 <= code AND code < 600,
+        total_count = COUNT(*)
+----
+
+Implementation wise, the base AggregateFunction has been extended to 
+allow a filter to be passed on. This is required to incorporate the
+filter as part of the aggregate equality/identity which would fail with 
+the filter as an external component.
+As part of the process, the serialization for the existing aggregations
+had to be fixed so AggregateFunction implementations so that it
+delegates to their parent first.
+
+{es-pull}113735[#113735]
+
 // end::notable-highlights[]
 
 
@@ -131,3 +156,44 @@ field is first combined into a single geometry before performing the predicate c
 
 {es-pull}112063[#112063]
 
+[discrete]
+[[enhance_sort_push_down_to_lucene_to_cover_references_to_fields_st_distance_function]]
+=== Enhance SORT push-down to Lucene to cover references to fields and ST_DISTANCE function
+The most used and likely most valuable geospatial search query in Elasticsearch is the sorted proximity search,
+finding items within a certain distance of a point of interest and sorting the results by distance.
+This has been possible in ES|QL since 8.15.0, but the sorting was done in-memory, not pushed down to Lucene.
+Now the sorting is pushed down to Lucene, which results in a significant performance improvement.
+
+Queries that perform both filtering and sorting on distance are supported. For example:
+
+[source,esql]
+----
+FROM test
+| EVAL distance = ST_DISTANCE(location, TO_GEOPOINT("POINT(37.7749, -122.4194)"))
+| WHERE distance < 1000000
+| SORT distance ASC, name DESC
+| LIMIT 10
+----
+
+In addition, the support for sorting on EVAL expressions has been extended to cover references to fields:
+
+[source,esql]
+----
+FROM test
+| EVAL ref = field
+| SORT ref ASC
+| LIMIT 10
+----
+
+{es-pull}112938[#112938]
+
+[discrete]
+[[cross_cluster_search_telemetry]]
+=== Cross-cluster search telemetry
+The cross-cluster search telemetry is collected when cross-cluster searches
+are performed, and is returned as "ccs" field in `_cluster/stats` output.
+It also add a new parameter `include_remotes=true` to the `_cluster/stats` API
+which will collect data from connected remote clusters.
+
+{es-pull}113825[#113825]
+