Add a direct IO option for rescoring to bbq_hnsw

qa/rolling-upgrade/src/javaRestTest/java/org/elasticsearch/upgrades/VectorSearchIT.java

@@ -16,6 +16,7 @@
 import org.elasticsearch.cluster.metadata.IndexMetadata;
 import org.elasticsearch.common.settings.Settings;
 import org.elasticsearch.common.xcontent.support.XContentMapValues;
+import org.elasticsearch.index.mapper.MapperFeatures;
 
 import java.io.IOException;
 import java.util.List;
@@ -35,6 +36,7 @@ public VectorSearchIT(@Name("upgradedNodes") int upgradedNodes) {
     private static final String BYTE_INDEX_NAME = "byte_vector_index";
     private static final String QUANTIZED_INDEX_NAME = "quantized_vector_index";
     private static final String BBQ_INDEX_NAME = "bbq_vector_index";
+    private static final String BBQ_INDEX_NAME_RESCORE = "bbq_vector_index_rescore";
     private static final String FLAT_QUANTIZED_INDEX_NAME = "flat_quantized_vector_index";
     private static final String FLAT_BBQ_INDEX_NAME = "flat_bbq_vector_index";
 
@@ -507,6 +509,63 @@ public void testBBQVectorSearch() throws Exception {
         );
     }
 
+    public void testBBQVectorSearchOffheapRescoring() throws Exception {
+        assumeTrue("Disabling off-heap rescoring is not supported", oldClusterHasFeature(MapperFeatures.BBQ_OFFHEAP_RESCORING));
+        if (isOldCluster()) {
+            String mapping = """
+                {
+                  "properties": {
+                    "vector": {
+                      "type": "dense_vector",
+                      "dims": 64,
+                      "index": true,
+                      "similarity": "cosine",
+                      "index_options": {
+                        "type": "bbq_hnsw",
+                        "ef_construction": 100,
+                        "m": 16,
+                        "disable_offheap_cache_rescoring": true
+                      }
+                    }
+                  }
+                }
+                """;
+            // create index and index 10 random floating point vectors
+            createIndex(
+                BBQ_INDEX_NAME_RESCORE,
+                Settings.builder().put(IndexMetadata.SETTING_NUMBER_OF_SHARDS, 1).put(IndexMetadata.SETTING_NUMBER_OF_REPLICAS, 0).build(),
+                mapping
+            );
+            index64DimVectors(BBQ_INDEX_NAME_RESCORE);
+            // force merge the index
+            client().performRequest(new Request("POST", "/" + BBQ_INDEX_NAME_RESCORE + "/_forcemerge?max_num_segments=1"));
+        }
+        Request searchRequest = new Request("POST", "/" + BBQ_INDEX_NAME_RESCORE + "/_search");
+        searchRequest.setJsonEntity("""
+            {
+                "knn": {
+                "field": "vector",
+                  "query_vector": [4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5,
+                   5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6],
+                  "k": 2,
+                  "num_candidates": 5,
+                  "rescore_vector": {
+                    "oversample": 2.0
+                  }
+                }
+            }
+            """);
+        Map<String, Object> response = search(searchRequest);
+        assertThat(extractValue(response, "hits.total.value"), equalTo(2));
+        List<Map<String, Object>> hits = extractValue(response, "hits.hits");
+        assertThat("expected: 0 received" + hits.get(0).get("_id") + " hits: " + response, hits.get(0).get("_id"), equalTo("0"));
+        assertThat(
+            "expected_near: 0.99 received" + hits.get(0).get("_score") + "hits: " + response,
+            (double) hits.get(0).get("_score"),
+            closeTo(0.9934857, 0.005)
+        );
+    }
+
     public void testFlatBBQVectorSearch() throws Exception {
         assumeTrue("Quantized vector search is not supported on this version", oldClusterHasFeature(BBQ_VECTOR_SEARCH_TEST_FEATURE));
         if (isOldCluster()) {

qa/vector/src/main/java/org/elasticsearch/test/knn/KnnIndexTester.java

@@ -106,7 +106,11 @@ private static String formatIndexPath(CmdLineArgs args) {
     static Codec createCodec(CmdLineArgs args) {
         final KnnVectorsFormat format;
         if (args.indexType() == IndexType.IVF) {
-            format = new ES920DiskBBQVectorsFormat(args.ivfClusterSize(), ES920DiskBBQVectorsFormat.DEFAULT_CENTROIDS_PER_PARENT_CLUSTER);
+            format = new ES920DiskBBQVectorsFormat(
+                args.ivfClusterSize(),
+                ES920DiskBBQVectorsFormat.DEFAULT_CENTROIDS_PER_PARENT_CLUSTER,
+                false
+            );
         } else {
             if (args.quantizeBits() == 1) {
                 if (args.indexType() == IndexType.FLAT) {

rest-api-spec/src/yamlRestTest/resources/rest-api-spec/test/search.vectors/41_knn_search_bbq_hnsw.yml

@@ -338,6 +338,69 @@ setup:
   - match: { hits.hits.1._score: $rescore_score1 }
   - match: { hits.hits.2._score: $rescore_score2 }
 ---
+"Test index configured rescore vector with no off-heap scoring":
+  - requires:
+      cluster_features: ["mapper.vectors.bbq_offheap_rescoring"]
+      reason: Needs bbq_offheap_rescoring feature
+  - skip:
+      features: "headers"
+  - do:
+      indices.create:
+        index: bbq_rescore_hnsw
+        body:
+          settings:
+            index:
+              number_of_shards: 1
+          mappings:
+            properties:
+              vector:
+                type: dense_vector
+                dims: 64
+                index: true
+                similarity: max_inner_product
+                index_options:
+                  type: bbq_hnsw
+                  disable_offheap_cache_rescoring: true
+                  rescore_vector:
+                    oversample: 1.5
+
+  - do:
+      bulk:
+        index: bbq_rescore_hnsw
+        refresh: true
+        body: |
+          { "index": {"_id": "1"}}
+          { "vector": [0.077,  0.32 , -0.205,  0.63 ,  0.032,  0.201,  0.167, -0.313, 0.176,  0.531, -0.375,  0.334, -0.046,  0.078, -0.349,  0.272, 0.307, -0.083,  0.504,  0.255, -0.404,  0.289, -0.226, -0.132, -0.216,  0.49 ,  0.039,  0.507, -0.307,  0.107,  0.09 , -0.265, -0.285,  0.336, -0.272,  0.369, -0.282,  0.086, -0.132,  0.475, -0.224,  0.203,  0.439,  0.064,  0.246, -0.396,  0.297,  0.242, -0.028,  0.321, -0.022, -0.009, -0.001  ,  0.031, -0.533,  0.45, -0.683,  1.331,  0.194, -0.157, -0.1  , -0.279, -0.098, -0.176] }
+          { "index": {"_id": "2"}}
+          { "vector": [0.196,  0.514,  0.039,  0.555, -0.042,  0.242,  0.463, -0.348, -0.08 ,  0.442, -0.067, -0.05 , -0.001,  0.298, -0.377,  0.048, 0.307,  0.159,  0.278,  0.119, -0.057,  0.333, -0.289, -0.438, -0.014,  0.361, -0.169,  0.292, -0.229,  0.123,  0.031, -0.138, -0.139,  0.315, -0.216,  0.322, -0.445, -0.059,  0.071,  0.429, -0.602, -0.142,  0.11 ,  0.192,  0.259, -0.241,  0.181, -0.166, 0.082,  0.107, -0.05 ,  0.155,  0.011,  0.161, -0.486,  0.569, -0.489,  0.901,  0.208,  0.011, -0.209, -0.153, -0.27 , -0.013] }
+          { "index": {"_id": "3"}}
+          { "vector": [0.196,  0.514,  0.039,  0.555, -0.042,  0.242,  0.463, -0.348, -0.08 ,  0.442, -0.067, -0.05 , -0.001,  0.298, -0.377,  0.048, 0.307,  0.159,  0.278,  0.119, -0.057,  0.333, -0.289, -0.438, -0.014,  0.361, -0.169,  0.292, -0.229,  0.123,  0.031, -0.138, -0.139,  0.315, -0.216,  0.322, -0.445, -0.059,  0.071,  0.429, -0.602, -0.142,  0.11 ,  0.192,  0.259, -0.241,  0.181, -0.166, 0.082,  0.107, -0.05 ,  0.155,  0.011,  0.161, -0.486,  0.569, -0.489,  0.901,  0.208,  0.011, -0.209, -0.153, -0.27 , -0.013] }
+
+  - do:
+      headers:
+        Content-Type: application/json
+      search:
+        rest_total_hits_as_int: true
+        index: bbq_rescore_hnsw
+        body:
+          knn:
+            field: vector
+            query_vector: [0.128,  0.067, -0.08 ,  0.395, -0.11 , -0.259,  0.473, -0.393,
+                           0.292,  0.571, -0.491,  0.444, -0.288,  0.198, -0.343,  0.015,
+                           0.232,  0.088,  0.228,  0.151, -0.136,  0.236, -0.273, -0.259,
+                           -0.217,  0.359, -0.207,  0.352, -0.142,  0.192, -0.061, -0.17 ,
+                           -0.343,  0.189, -0.221,  0.32 , -0.301, -0.1  ,  0.005,  0.232,
+                           -0.344,  0.136,  0.252,  0.157, -0.13 , -0.244,  0.193, -0.034,
+                           -0.12 , -0.193, -0.102,  0.252, -0.185, -0.167, -0.575,  0.582,
+                           -0.426,  0.983,  0.212,  0.204,  0.03 , -0.276, -0.425, -0.158]
+            k: 3
+            num_candidates: 3
+
+  - match: { hits.total: 3 }
+  - set: { hits.hits.0._score: rescore_score0 }
+  - set: { hits.hits.1._score: rescore_score1 }
+  - set: { hits.hits.2._score: rescore_score2 }
+---
 "Test index configured rescore vector updateable and settable to 0":
   - requires:
       cluster_features: ["mapper.dense_vector.rescore_zero_vector"]

server/src/internalClusterTest/java/org/elasticsearch/index/store/DirectIOIT.java

@@ -17,7 +17,6 @@
 import org.apache.lucene.store.IndexOutput;
 import org.apache.lucene.tests.util.LuceneTestCase;
 import org.elasticsearch.common.settings.Settings;
-import org.elasticsearch.index.codec.vectors.es818.ES818BinaryQuantizedVectorsFormat;
 import org.elasticsearch.plugins.Plugin;
 import org.elasticsearch.search.vectors.KnnSearchBuilder;
 import org.elasticsearch.search.vectors.VectorData;
@@ -50,8 +49,6 @@ public class DirectIOIT extends ESIntegTestCase {
 
     @BeforeClass
     public static void checkSupported() {
-        assumeTrue("Direct IO is not enabled", ES818BinaryQuantizedVectorsFormat.USE_DIRECT_IO);
-
         Path path = createTempDir("directIOProbe");
         try (Directory dir = open(path); IndexOutput out = dir.createOutput("out", IOContext.DEFAULT)) {
             out.writeString("test");
@@ -76,7 +73,7 @@ protected Collection<Class<? extends Plugin>> nodePlugins() {
     }
 
     private void indexVectors() {
-        String type = randomFrom("bbq_flat", "bbq_hnsw");
+        String type = "bbq_hnsw";
         assertAcked(
             prepareCreate("foo-vectors").setSettings(Settings.builder().put(InternalSettingsPlugin.USE_COMPOUND_FILE.getKey(), false))
                 .setMapping("""
@@ -89,7 +86,8 @@ private void indexVectors() {
                           "index": true,
                           "similarity": "l2_norm",
                           "index_options": {
-                            "type": "%type%"
+                            "type": "%type%",
+                            "disable_offheap_cache_rescoring": true
                           }
                         }
                       }
@@ -105,11 +103,12 @@ private void indexVectors() {
         assertBBQIndexType(type); // test assertion to ensure that the correct index type is being used
     }
 
-    @SuppressWarnings("unchecked")
     static void assertBBQIndexType(String type) {
         var response = indicesAdmin().prepareGetFieldMappings("foo-vectors").setFields("fooVector").get();
-        var map = (Map<String, Object>) response.fieldMappings("foo-vectors", "fooVector").sourceAsMap().get("fooVector");
-        assertThat((String) ((Map<String, Object>) map.get("index_options")).get("type"), is(equalTo(type)));
+        var map = (Map<?, ?>) response.fieldMappings("foo-vectors", "fooVector").sourceAsMap().get("fooVector");
+        var options = (Map<?, ?>) map.get("index_options");
+        assertThat(options.get("type"), is(equalTo(type)));
+        assertThat(options.get("disable_offheap_cache_rescoring"), is(true));
     }
 
     @TestLogging(value = "org.elasticsearch.index.store.FsDirectoryFactory:DEBUG", reason = "to capture trace logging for direct IO")

server/src/main/java/module-info.java

@@ -7,7 +7,6 @@
  * License v3.0 only", or the "Server Side Public License, v 1".
  */
 
-import org.elasticsearch.index.codec.vectors.diskbbq.ES920DiskBBQVectorsFormat;
 import org.elasticsearch.plugins.internal.RestExtension;
 import org.elasticsearch.reservedstate.ReservedStateHandlerProvider;
 
@@ -462,7 +461,7 @@
             org.elasticsearch.index.codec.vectors.es816.ES816HnswBinaryQuantizedVectorsFormat,
             org.elasticsearch.index.codec.vectors.es818.ES818BinaryQuantizedVectorsFormat,
             org.elasticsearch.index.codec.vectors.es818.ES818HnswBinaryQuantizedVectorsFormat,
-            ES920DiskBBQVectorsFormat;
+            org.elasticsearch.index.codec.vectors.diskbbq.ES920DiskBBQVectorsFormat;
 
     provides org.apache.lucene.codecs.Codec
         with

server/src/main/java/org/elasticsearch/index/codec/vectors/AbstractFlatVectorsFormat.java

@@ -11,21 +11,11 @@
 
 import org.apache.lucene.codecs.hnsw.FlatVectorsFormat;
 import org.apache.lucene.codecs.hnsw.FlatVectorsScorer;
-import org.elasticsearch.core.SuppressForbidden;
 
 import static org.elasticsearch.index.mapper.vectors.DenseVectorFieldMapper.MAX_DIMS_COUNT;
 
 public abstract class AbstractFlatVectorsFormat extends FlatVectorsFormat {
 
-    public static final boolean USE_DIRECT_IO = getUseDirectIO();
-
-    @SuppressForbidden(
-        reason = "TODO Deprecate any lenient usage of Boolean#parseBoolean https://github.com/elastic/elasticsearch/issues/128993"
-    )
-    private static boolean getUseDirectIO() {
-        return Boolean.parseBoolean(System.getProperty("vector.rescoring.directio", "false"));
-    }
-
     protected AbstractFlatVectorsFormat(String name) {
         super(name);
     }

server/src/main/java/org/elasticsearch/index/codec/vectors/diskbbq/ES920DiskBBQVectorsFormat.java

@@ -20,6 +20,7 @@
 import org.apache.lucene.index.SegmentWriteState;
 import org.elasticsearch.common.util.Maps;
 import org.elasticsearch.index.codec.vectors.OptimizedScalarQuantizer;
+import org.elasticsearch.index.codec.vectors.es818.DirectIOLucene99FlatVectorsFormat;
 
 import java.io.IOException;
 import java.util.Collections;
@@ -63,7 +64,15 @@ public class ES920DiskBBQVectorsFormat extends KnnVectorsFormat {
     private static final FlatVectorsFormat rawVectorFormat = new Lucene99FlatVectorsFormat(
         FlatVectorScorerUtil.getLucene99FlatVectorsScorer()
     );
-    private static final Map<String, FlatVectorsFormat> supportedFormats = Map.of(rawVectorFormat.getName(), rawVectorFormat);
+    private static final FlatVectorsFormat directIORawVectorFormat = new DirectIOLucene99FlatVectorsFormat(
+        FlatVectorScorerUtil.getLucene99FlatVectorsScorer()
+    );
+    private static final Map<String, FlatVectorsFormat> supportedFormats = Map.of(
+        rawVectorFormat.getName(),
+        rawVectorFormat,
+        directIORawVectorFormat.getName(),
+        directIORawVectorFormat
+    );
 
     // This dynamically sets the cluster probe based on the `k` requested and the number of clusters.
     // useful when searching with 'efSearch' type parameters instead of requiring a specific ratio.
@@ -77,8 +86,9 @@ public class ES920DiskBBQVectorsFormat extends KnnVectorsFormat {
 
     private final int vectorPerCluster;
     private final int centroidsPerParentCluster;
+    private final boolean directRawDiskReads;
 
-    public ES920DiskBBQVectorsFormat(int vectorPerCluster, int centroidsPerParentCluster) {
+    public ES920DiskBBQVectorsFormat(int vectorPerCluster, int centroidsPerParentCluster, boolean directRawDiskReads) {
         super(NAME);
         if (vectorPerCluster < MIN_VECTORS_PER_CLUSTER || vectorPerCluster > MAX_VECTORS_PER_CLUSTER) {
             throw new IllegalArgumentException(
@@ -102,19 +112,21 @@ public ES920DiskBBQVectorsFormat(int vectorPerCluster, int centroidsPerParentClu
         }
         this.vectorPerCluster = vectorPerCluster;
         this.centroidsPerParentCluster = centroidsPerParentCluster;
+        this.directRawDiskReads = directRawDiskReads;
     }
 
     /** Constructs a format using the given graph construction parameters and scalar quantization. */
     public ES920DiskBBQVectorsFormat() {
-        this(DEFAULT_VECTORS_PER_CLUSTER, DEFAULT_CENTROIDS_PER_PARENT_CLUSTER);
+        this(DEFAULT_VECTORS_PER_CLUSTER, DEFAULT_CENTROIDS_PER_PARENT_CLUSTER, false);
     }
 
     @Override
     public KnnVectorsWriter fieldsWriter(SegmentWriteState state) throws IOException {
+        FlatVectorsFormat rawFormat = directRawDiskReads ? directIORawVectorFormat : rawVectorFormat;
         return new ES920DiskBBQVectorsWriter(
-            rawVectorFormat.getName(),
+            rawFormat.getName(),
             state,
-            rawVectorFormat.fieldsWriter(state),
+            rawFormat.fieldsWriter(state),
             vectorPerCluster,
             centroidsPerParentCluster
         );

server/src/main/java/org/elasticsearch/index/codec/vectors/es818/DirectIOLucene99FlatVectorsFormat.java

@@ -41,11 +41,10 @@
  * Copied from Lucene99FlatVectorsFormat in Lucene 10.1
  *
  * This is copied to change the implementation of {@link #fieldsReader} only.
- * The codec format itself is not changed, so we keep the original {@link #NAME}
  */
 public class DirectIOLucene99FlatVectorsFormat extends AbstractFlatVectorsFormat {
 
-    static final String NAME = "Lucene99FlatVectorsFormat";
+    static final String NAME = "DirectIOLucene99FlatVectorsFormat";
 
     public static final int VERSION_START = 0;
     public static final int VERSION_CURRENT = VERSION_START;
@@ -69,7 +68,6 @@ public FlatVectorsWriter fieldsWriter(SegmentWriteState state) throws IOExceptio
     }
 
     static boolean shouldUseDirectIO(SegmentReadState state) {
-        assert USE_DIRECT_IO;
         return FsDirectoryFactory.isHybridFs(state.directory);
     }
 

server/src/main/java/org/elasticsearch/index/codec/vectors/es818/ES818BinaryQuantizedVectorsFormat.java

@@ -98,17 +98,20 @@ public class ES818BinaryQuantizedVectorsFormat extends AbstractFlatVectorsFormat
     static final String VECTOR_DATA_EXTENSION = "veb";
     static final int DIRECT_MONOTONIC_BLOCK_SHIFT = 16;
 
-    private static final FlatVectorsFormat rawVectorFormat = USE_DIRECT_IO
-        ? new DirectIOLucene99FlatVectorsFormat(FlatVectorScorerUtil.getLucene99FlatVectorsScorer())
-        : new Lucene99FlatVectorsFormat(FlatVectorScorerUtil.getLucene99FlatVectorsScorer());
-
-    private static final ES818BinaryFlatVectorsScorer scorer = new ES818BinaryFlatVectorsScorer(
+    static final ES818BinaryFlatVectorsScorer scorer = new ES818BinaryFlatVectorsScorer(
         FlatVectorScorerUtil.getLucene99FlatVectorsScorer()
     );
 
+    private final FlatVectorsFormat rawVectorFormat;
+
     /** Creates a new instance with the default number of vectors per cluster. */
     public ES818BinaryQuantizedVectorsFormat() {
-        super(NAME);
+        this(NAME, new Lucene99FlatVectorsFormat(FlatVectorScorerUtil.getLucene99FlatVectorsScorer()));
+    }
+
+    ES818BinaryQuantizedVectorsFormat(String name, FlatVectorsFormat rawVectorFormat) {
+        super(name);
+        this.rawVectorFormat = rawVectorFormat;
     }
 
     @Override

server/src/main/java/org/elasticsearch/index/mapper/MapperFeatures.java

@@ -51,6 +51,7 @@ public class MapperFeatures implements FeatureSpecification {
     static final NodeFeature PATTERN_TEXT = new NodeFeature("mapper.patterned_text");
     static final NodeFeature IGNORED_SOURCE_FIELDS_PER_ENTRY = new NodeFeature("mapper.ignored_source_fields_per_entry");
     public static final NodeFeature MULTI_FIELD_UNICODE_OPTIMISATION_FIX = new NodeFeature("mapper.multi_field.unicode_optimisation_fix");
+    public static final NodeFeature BBQ_OFFHEAP_RESCORING = new NodeFeature("mapper.vectors.bbq_offheap_rescoring");
 
     @Override
     public Set<NodeFeature> getTestFeatures() {
@@ -87,7 +88,8 @@ public Set<NodeFeature> getTestFeatures() {
             PATTERN_TEXT,
             IGNORED_SOURCE_FIELDS_PER_ENTRY,
             MULTI_FIELD_UNICODE_OPTIMISATION_FIX,
-            MATCH_ONLY_TEXT_BLOCK_LOADER_FIX
+            MATCH_ONLY_TEXT_BLOCK_LOADER_FIX,
+            BBQ_OFFHEAP_RESCORING
         );
     }
 }