Add dense vector off-heap stats to Node stats and Index stats APIs (#126704)

This change enhances the dense_vector section of the Nodes stats and Index stats APIs so that they report the desired size of off-heap memory for all indexed vectors. The dense_vector section of the Custer stats API remains unchanged.

The retrieval mechanism and structure of the new stats is the same across the various three stats APIs, but more fine-grained information is disclosed as when moving from Cluster -> Node -> Index API.

For Node stats, we aggregate the total byte sizes for all vectors, categorised by the data type. For example:

"dense_vector" : {
  "value_count" : 5,
  "off_heap" : {
    "total_size_in_bytes" : 27,
    "total_veb_size_in_bytes" : 3,
    "total_vec_size_in_bytes" : 23,
    "total_veq_size_in_bytes" : 0,
    "total_vex_size_in_bytes" : 1
  }
}
Index stats: same as Node stats with included field break down . For example:

"dense_vector" : {
  "value_count" : 5,
  "off_heap" : {
    "total_size_in_bytes" : 27,
    "total_veb_size_in_bytes" : 3,
    "total_vec_size_in_bytes" : 23,
    "total_veq_size_in_bytes" : 0,
    "total_vex_size_in_bytes" : 1,
    "fielddata" : {
      "bar" : {
        "veb_size_in_bytes" : 3,
        "vec_size_in_bytes" : 14,
        "vex_size_in_bytes" : 1
      },
      "foo" : {
        "vec_size_in_bytes" : 9
      }
    }
  }
The implementation accesses the actual statistics through reflection. This will be completely removed when Lucene exposes this, which is expected in Lucene 10.3

Author: ChrisHegarty

distribution/tools/server-cli/src/main/java/org/elasticsearch/server/cli/SystemJvmOptions.java

@@ -61,6 +61,13 @@ static List<String> systemJvmOptions(Settings nodeSettings, final Map<String, St
                 "-Dio.netty.noUnsafe=true",
                 "-Dio.netty.noKeySetOptimization=true",
                 "-Dio.netty.recycler.maxCapacityPerThread=0",
+                // temporary until we get off-heap vector stats in Lucene 10.3
+                "--add-opens=org.apache.lucene.core/org.apache.lucene.codecs.lucene99=org.elasticsearch.server",
+                "--add-opens=org.apache.lucene.backward_codecs/org.apache.lucene.backward_codecs.lucene90=org.elasticsearch.server",
+                "--add-opens=org.apache.lucene.backward_codecs/org.apache.lucene.backward_codecs.lucene91=org.elasticsearch.server",
+                "--add-opens=org.apache.lucene.backward_codecs/org.apache.lucene.backward_codecs.lucene92=org.elasticsearch.server",
+                "--add-opens=org.apache.lucene.backward_codecs/org.apache.lucene.backward_codecs.lucene94=org.elasticsearch.server",
+                "--add-opens=org.apache.lucene.backward_codecs/org.apache.lucene.backward_codecs.lucene95=org.elasticsearch.server",
                 // log4j 2
                 "-Dlog4j.shutdownHookEnabled=false",
                 "-Dlog4j2.disable.jmx=true",

docs/changelog/126704.yaml

@@ -0,0 +1,5 @@
+pr: 126704
+summary: Add dense vector off-heap stats to Node stats and Index stats APIs
+area: "Vector Search"
+type: enhancement
+issues: []

qa/ccs-common-rest/build.gradle

@@ -10,8 +10,8 @@ apply plugin: 'elasticsearch.internal-yaml-rest-test'
 
 restResources {
   restApi {
-    include 'capabilities', '_common', 'bulk', 'count', 'cluster', 'field_caps', 'get', 'knn_search', 'index', 'indices', 'msearch',
-      'search', 'async_search', 'graph', '*_point_in_time', 'info', 'scroll', 'clear_scroll', 'search_mvt', 'eql', 'sql'
+    include 'capabilities', 'cat.shards', '_common', 'bulk', 'count', 'cluster', 'field_caps', 'get', 'knn_search', 'index', 'indices', 'msearch',
+      "nodes.stats", 'search', 'async_search', 'graph', '*_point_in_time', 'info', 'scroll', 'clear_scroll', 'search_mvt', 'eql', 'sql'
   }
   restTests {
     includeCore 'field_caps', 'msearch', 'search', 'suggest', 'scroll', "indices.resolve_index"

rest-api-spec/src/yamlRestTest/resources/rest-api-spec/test/cluster.stats/10_basic.yml

@@ -351,6 +351,7 @@
   - match: { indices.docs.count: 5 }
   - match: { indices.docs.deleted: 0 }
   - match: { indices.dense_vector.value_count: 8 }
+  - is_false: indices.dense_vector.off_heap
 
 ---
 "Sparse vector stats":