Merge branch 'main' into modified-flat-vector-direct-io

Author: thecoop

.buildkite/scripts/dra-workflow.sh

@@ -70,6 +70,7 @@ echo --- Building release artifacts
   $VERSION_QUALIFIER_ARG \
   buildReleaseArtifacts \
   exportCompressedDockerImages \
+  exportDockerContexts \
   :distribution:generateDependenciesReport
 
 PATH="$PATH:${JAVA_HOME}/bin" # Required by the following script

.ci/scripts/packaging-test.sh

@@ -3,7 +3,7 @@
 # opensuse 15 has a missing dep for systemd
 
 if which zypper > /dev/null ; then
-    sudo zypper install -y insserv-compat
+    sudo zypper install -y insserv-compat docker-buildx
 fi
 
 if [ -e /etc/sysctl.d/99-gce.conf ]; then

benchmarks/src/main/java/org/elasticsearch/benchmark/index/codec/tsdb/TSDBDocValuesMergeBenchmark.java

@@ -0,0 +1,196 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the "Elastic License
+ * 2.0", the "GNU Affero General Public License v3.0 only", and the "Server Side
+ * Public License v 1"; you may not use this file except in compliance with, at
+ * your election, the "Elastic License 2.0", the "GNU Affero General Public
+ * License v3.0 only", or the "Server Side Public License, v 1".
+ */
+
+package org.elasticsearch.benchmark.index.codec.tsdb;
+
+import org.apache.lucene.analysis.standard.StandardAnalyzer;
+import org.apache.lucene.codecs.DocValuesFormat;
+import org.apache.lucene.document.Document;
+import org.apache.lucene.document.SortedDocValuesField;
+import org.apache.lucene.document.SortedNumericDocValuesField;
+import org.apache.lucene.document.SortedSetDocValuesField;
+import org.apache.lucene.index.IndexWriter;
+import org.apache.lucene.index.IndexWriterConfig;
+import org.apache.lucene.index.LogByteSizeMergePolicy;
+import org.apache.lucene.search.Sort;
+import org.apache.lucene.search.SortField;
+import org.apache.lucene.search.SortedNumericSortField;
+import org.apache.lucene.store.Directory;
+import org.apache.lucene.store.FSDirectory;
+import org.apache.lucene.util.BytesRef;
+import org.elasticsearch.cluster.metadata.DataStream;
+import org.elasticsearch.common.logging.LogConfigurator;
+import org.elasticsearch.index.codec.Elasticsearch900Lucene101Codec;
+import org.elasticsearch.index.codec.tsdb.es819.ES819TSDBDocValuesFormat;
+import org.openjdk.jmh.annotations.Benchmark;
+import org.openjdk.jmh.annotations.BenchmarkMode;
+import org.openjdk.jmh.annotations.Fork;
+import org.openjdk.jmh.annotations.Level;
+import org.openjdk.jmh.annotations.Measurement;
+import org.openjdk.jmh.annotations.Mode;
+import org.openjdk.jmh.annotations.OutputTimeUnit;
+import org.openjdk.jmh.annotations.Param;
+import org.openjdk.jmh.annotations.Scope;
+import org.openjdk.jmh.annotations.Setup;
+import org.openjdk.jmh.annotations.State;
+import org.openjdk.jmh.annotations.TearDown;
+import org.openjdk.jmh.annotations.Threads;
+import org.openjdk.jmh.annotations.Warmup;
+import org.openjdk.jmh.profile.AsyncProfiler;
+import org.openjdk.jmh.runner.Runner;
+import org.openjdk.jmh.runner.RunnerException;
+import org.openjdk.jmh.runner.options.Options;
+import org.openjdk.jmh.runner.options.OptionsBuilder;
+
+import java.io.IOException;
+import java.nio.file.Files;
+import java.util.Random;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.TimeUnit;
+
+@BenchmarkMode(Mode.SampleTime)
+@OutputTimeUnit(TimeUnit.MILLISECONDS)
+@State(Scope.Benchmark)
+@Fork(1)
+@Threads(1)
+@Warmup(iterations = 0)
+@Measurement(iterations = 1)
+public class TSDBDocValuesMergeBenchmark {
+
+    static {
+        // For Elasticsearch900Lucene101Codec:
+        LogConfigurator.loadLog4jPlugins();
+        LogConfigurator.configureESLogging();
+        LogConfigurator.setNodeName("test");
+    }
+
+    @Param("20431204")
+    private int nDocs;
+
+    @Param("1000")
+    private int deltaTime;
+
+    @Param("42")
+    private int seed;
+
+    private static final String TIMESTAMP_FIELD = "@timestamp";
+    private static final String HOSTNAME_FIELD = "host.name";
+    private static final long BASE_TIMESTAMP = 1704067200000L;
+
+    private IndexWriter indexWriterWithoutOptimizedMerge;
+    private IndexWriter indexWriterWithOptimizedMerge;
+    private ExecutorService executorService;
+
+    public static void main(String[] args) throws RunnerException {
+        final Options options = new OptionsBuilder().include(TSDBDocValuesMergeBenchmark.class.getSimpleName())
+            .addProfiler(AsyncProfiler.class)
+            .build();
+
+        new Runner(options).run();
+    }
+
+    @Setup(Level.Trial)
+    public void setup() throws IOException {
+        executorService = Executors.newSingleThreadExecutor();
+
+        final Directory tempDirectoryWithoutDocValuesSkipper = FSDirectory.open(Files.createTempDirectory("temp1-"));
+        final Directory tempDirectoryWithDocValuesSkipper = FSDirectory.open(Files.createTempDirectory("temp2-"));
+
+        indexWriterWithoutOptimizedMerge = createIndex(tempDirectoryWithoutDocValuesSkipper, false);
+        indexWriterWithOptimizedMerge = createIndex(tempDirectoryWithDocValuesSkipper, true);
+    }
+
+    private IndexWriter createIndex(final Directory directory, final boolean optimizedMergeEnabled) throws IOException {
+        final var iwc = createIndexWriterConfig(optimizedMergeEnabled);
+        long counter1 = 0;
+        long counter2 = 10_000_000;
+        long[] gauge1Values = new long[] { 2, 4, 6, 8, 10, 12, 14, 16 };
+        long[] gauge2Values = new long[] { -2, -4, -6, -8, -10, -12, -14, -16 };
+        int numHosts = 1000;
+        String[] tags = new String[] { "tag_1", "tag_2", "tag_3", "tag_4", "tag_5", "tag_6", "tag_7", "tag_8" };
+
+        final Random random = new Random(seed);
+        IndexWriter indexWriter = new IndexWriter(directory, iwc);
+        for (int i = 0; i < nDocs; i++) {
+            final Document doc = new Document();
+
+            final int batchIndex = i / numHosts;
+            final String hostName = "host-" + batchIndex;
+            // Slightly vary the timestamp in each document
+            final long timestamp = BASE_TIMESTAMP + ((i % numHosts) * deltaTime) + random.nextInt(0, deltaTime);
+
+            doc.add(new SortedDocValuesField(HOSTNAME_FIELD, new BytesRef(hostName)));
+            doc.add(new SortedNumericDocValuesField(TIMESTAMP_FIELD, timestamp));
+            doc.add(new SortedNumericDocValuesField("counter_1", counter1++));
+            doc.add(new SortedNumericDocValuesField("counter_2", counter2++));
+            doc.add(new SortedNumericDocValuesField("gauge_1", gauge1Values[i % gauge1Values.length]));
+            doc.add(new SortedNumericDocValuesField("gauge_2", gauge2Values[i % gauge1Values.length]));
+            int numTags = tags.length % (i + 1);
+            for (int j = 0; j < numTags; j++) {
+                doc.add(new SortedSetDocValuesField("tags", new BytesRef(tags[j])));
+            }
+
+            indexWriter.addDocument(doc);
+        }
+        indexWriter.commit();
+        return indexWriter;
+    }
+
+    @Benchmark
+    public void forceMergeWithoutOptimizedMerge() throws IOException {
+        forceMerge(indexWriterWithoutOptimizedMerge);
+    }
+
+    @Benchmark
+    public void forceMergeWithOptimizedMerge() throws IOException {
+        forceMerge(indexWriterWithOptimizedMerge);
+    }
+
+    private void forceMerge(final IndexWriter indexWriter) throws IOException {
+        indexWriter.forceMerge(1);
+    }
+
+    @TearDown(Level.Trial)
+    public void tearDown() {
+        if (executorService != null) {
+            executorService.shutdown();
+            try {
+                if (executorService.awaitTermination(30, TimeUnit.SECONDS) == false) {
+                    executorService.shutdownNow();
+                }
+            } catch (InterruptedException e) {
+                executorService.shutdownNow();
+                Thread.currentThread().interrupt();
+            }
+        }
+    }
+
+    private static IndexWriterConfig createIndexWriterConfig(boolean optimizedMergeEnabled) {
+        var config = new IndexWriterConfig(new StandardAnalyzer());
+        // NOTE: index sort config matching LogsDB's sort order
+        config.setIndexSort(
+            new Sort(
+                new SortField(HOSTNAME_FIELD, SortField.Type.STRING, false),
+                new SortedNumericSortField(TIMESTAMP_FIELD, SortField.Type.LONG, true)
+            )
+        );
+        config.setLeafSorter(DataStream.TIMESERIES_LEAF_READERS_SORTER);
+        config.setMergePolicy(new LogByteSizeMergePolicy());
+        var docValuesFormat = new ES819TSDBDocValuesFormat(4096, optimizedMergeEnabled);
+        config.setCodec(new Elasticsearch900Lucene101Codec() {
+
+            @Override
+            public DocValuesFormat getDocValuesFormatForField(String field) {
+                return docValuesFormat;
+            }
+        });
+        return config;
+    }
+}

benchmarks/src/main/java/org/elasticsearch/benchmark/routing/allocation/AllocationBenchmark.java

@@ -14,7 +14,10 @@
 import org.elasticsearch.cluster.ClusterState;
 import org.elasticsearch.cluster.metadata.IndexMetadata;
 import org.elasticsearch.cluster.metadata.Metadata;
+import org.elasticsearch.cluster.metadata.ProjectId;
+import org.elasticsearch.cluster.metadata.ProjectMetadata;
 import org.elasticsearch.cluster.node.DiscoveryNodes;
+import org.elasticsearch.cluster.routing.GlobalRoutingTable;
 import org.elasticsearch.cluster.routing.RoutingTable;
 import org.elasticsearch.cluster.routing.ShardRouting;
 import org.elasticsearch.cluster.routing.allocation.AllocationService;
@@ -126,19 +129,20 @@ public void setUp() throws Exception {
             Settings.builder().put("cluster.routing.allocation.awareness.attributes", "tag").build()
         );
 
-        Metadata.Builder mb = Metadata.builder();
+        final ProjectId projectId = ProjectId.DEFAULT;
+        ProjectMetadata.Builder pmb = ProjectMetadata.builder(projectId);
         for (int i = 1; i <= numIndices; i++) {
-            mb.put(
+            pmb.put(
                 IndexMetadata.builder("test_" + i)
                     .settings(Settings.builder().put(IndexMetadata.SETTING_VERSION_CREATED, IndexVersion.current()))
                     .numberOfShards(numShards)
                     .numberOfReplicas(numReplicas)
             );
         }
-        Metadata metadata = mb.build();
+        Metadata metadata = Metadata.builder().put(pmb).build();
         RoutingTable.Builder rb = RoutingTable.builder(TestShardRoutingRoleStrategies.DEFAULT_ROLE_ONLY);
         for (int i = 1; i <= numIndices; i++) {
-            rb.addAsNew(metadata.getProject().index("test_" + i));
+            rb.addAsNew(metadata.getProject(projectId).index("test_" + i));
         }
         RoutingTable routingTable = rb.build();
         DiscoveryNodes.Builder nb = DiscoveryNodes.builder();
@@ -151,7 +155,7 @@ public void setUp() throws Exception {
         }
         initialClusterState = ClusterState.builder(ClusterName.DEFAULT)
             .metadata(metadata)
-            .routingTable(routingTable)
+            .routingTable(GlobalRoutingTable.builder().put(projectId, routingTable).build())
             .nodes(nb)
             .nodeIdsToCompatibilityVersions(compatibilityVersions)
             .build();

benchmarks/src/main/java/org/elasticsearch/benchmark/vector/VectorScorerBenchmark.java

@@ -19,6 +19,7 @@
 import org.apache.lucene.store.MMapDirectory;
 import org.apache.lucene.util.hnsw.RandomVectorScorer;
 import org.apache.lucene.util.hnsw.RandomVectorScorerSupplier;
+import org.apache.lucene.util.hnsw.UpdateableRandomVectorScorer;
 import org.apache.lucene.util.quantization.QuantizedByteVectorValues;
 import org.apache.lucene.util.quantization.ScalarQuantizer;
 import org.elasticsearch.common.logging.LogConfigurator;
@@ -76,10 +77,10 @@ public class VectorScorerBenchmark {
     float vec2Offset;
     float scoreCorrectionConstant;
 
-    RandomVectorScorer luceneDotScorer;
-    RandomVectorScorer luceneSqrScorer;
-    RandomVectorScorer nativeDotScorer;
-    RandomVectorScorer nativeSqrScorer;
+    UpdateableRandomVectorScorer luceneDotScorer;
+    UpdateableRandomVectorScorer luceneSqrScorer;
+    UpdateableRandomVectorScorer nativeDotScorer;
+    UpdateableRandomVectorScorer nativeSqrScorer;
 
     RandomVectorScorer luceneDotScorerQuery;
     RandomVectorScorer nativeDotScorerQuery;
@@ -118,12 +119,16 @@ public void setup() throws IOException {
         in = dir.openInput("vector.data", IOContext.DEFAULT);
         var values = vectorValues(dims, 2, in, VectorSimilarityFunction.DOT_PRODUCT);
         scoreCorrectionConstant = values.getScalarQuantizer().getConstantMultiplier();
-        luceneDotScorer = luceneScoreSupplier(values, VectorSimilarityFunction.DOT_PRODUCT).scorer(0);
+        luceneDotScorer = luceneScoreSupplier(values, VectorSimilarityFunction.DOT_PRODUCT).scorer();
+        luceneDotScorer.setScoringOrdinal(0);
         values = vectorValues(dims, 2, in, VectorSimilarityFunction.EUCLIDEAN);
-        luceneSqrScorer = luceneScoreSupplier(values, VectorSimilarityFunction.EUCLIDEAN).scorer(0);
+        luceneSqrScorer = luceneScoreSupplier(values, VectorSimilarityFunction.EUCLIDEAN).scorer();
+        luceneSqrScorer.setScoringOrdinal(0);
 
-        nativeDotScorer = factory.getInt7SQVectorScorerSupplier(DOT_PRODUCT, in, values, scoreCorrectionConstant).get().scorer(0);
-        nativeSqrScorer = factory.getInt7SQVectorScorerSupplier(EUCLIDEAN, in, values, scoreCorrectionConstant).get().scorer(0);
+        nativeDotScorer = factory.getInt7SQVectorScorerSupplier(DOT_PRODUCT, in, values, scoreCorrectionConstant).get().scorer();
+        nativeDotScorer.setScoringOrdinal(0);
+        nativeSqrScorer = factory.getInt7SQVectorScorerSupplier(EUCLIDEAN, in, values, scoreCorrectionConstant).get().scorer();
+        nativeSqrScorer.setScoringOrdinal(0);
 
         // setup for getInt7SQVectorScorer / query vector scoring
         float[] queryVec = new float[dims];

build-tools-internal/src/integTest/groovy/org/elasticsearch/gradle/internal/test/rest/LegacyYamlRestTestPluginFuncTest.groovy

@@ -179,9 +179,9 @@ echo "Running elasticsearch \$0"
 
         file(distProjectFolder, 'src/config/elasticsearch.properties') << "some propes"
         file(distProjectFolder, 'src/config/jvm.options') << """
--Xlog:gc*,gc+age=trace,safepoint:file=logs/gc.log:utctime,level,pid,tags:filecount=32,filesize=64m
--XX:ErrorFile=logs/hs_err_pid%p.log
--XX:HeapDumpPath=data
+-Xlog:gc*,gc+age=trace,safepoint:file=gc.log:utctime,level,pid,tags:filecount=32,filesize=64m
+-XX:ErrorFile=hs_err_pid%p.log
+# -XX:HeapDumpPath=/heap/dump/path
 """
         file(distProjectFolder, 'build.gradle') << """
             import org.gradle.api.internal.artifacts.ArtifactAttributes;

build-tools-internal/src/main/java/org/elasticsearch/gradle/internal/DockerBase.java

@@ -14,36 +14,45 @@
  */
 public enum DockerBase {
     // "latest" here is intentional, since the image name specifies "9"
-    DEFAULT("redhat/ubi9-minimal:latest", "", "microdnf"),
+    DEFAULT("redhat/ubi9-minimal:latest", "", "microdnf", "Dockerfile.default"),
 
     // The Iron Bank base image is UBI (albeit hardened), but we are required to parameterize the Docker build
-    IRON_BANK("${BASE_REGISTRY}/${BASE_IMAGE}:${BASE_TAG}", "-ironbank", "yum"),
+    IRON_BANK("${BASE_REGISTRY}/${BASE_IMAGE}:${BASE_TAG}", "-ironbank", "yum", "Dockerfile"),
 
     // Chainguard based wolfi image with latest jdk
     // This is usually updated via renovatebot
     // spotless:off
-    WOLFI("docker.elastic.co/wolfi/chainguard-base:latest@sha256:29150cd940cc7f69407d978d5a19c86f4d9e67cf44e4d6ded787a497e8f27c9a",
+    WOLFI(
+        "docker.elastic.co/wolfi/chainguard-base:latest@sha256:29150cd940cc7f69407d978d5a19c86f4d9e67cf44e4d6ded787a497e8f27c9a",
         "-wolfi",
-        "apk"
+        "apk",
+        "Dockerfile"
+    ),
+    FIPS(
+        "docker.elastic.co/wolfi/chainguard-base-fips:sha256-ebfc3f1d7dba992231747a2e05ad1b859843e81b5e676ad342859d7cf9e425a7@sha256:ebfc3f1d7dba992231747a2e05ad1b859843e81b5e676ad342859d7cf9e425a7",
+        "-fips",
+        "apk",
+        "Dockerfile"
     ),
-    FIPS("docker.elastic.co/wolfi/chainguard-base-fips:sha256-ebfc3f1d7dba992231747a2e05ad1b859843e81b5e676ad342859d7cf9e425a7@sha256:ebfc3f1d7dba992231747a2e05ad1b859843e81b5e676ad342859d7cf9e425a7", "-fips", "apk"),
     // spotless:on
     // Based on WOLFI above, with more extras. We don't set a base image because
     // we programmatically extend from the wolfi image.
-    CLOUD_ESS(null, "-cloud-ess", "apk");
+    CLOUD_ESS(null, "-cloud-ess", "apk", "Dockerfile.cloud-ess"),;
 
     private final String image;
     private final String suffix;
     private final String packageManager;
+    private final String dockerfile;
 
     DockerBase(String image, String suffix) {
-        this(image, suffix, "apt-get");
+        this(image, suffix, "apt-get", "dockerfile");
     }
 
-    DockerBase(String image, String suffix, String packageManager) {
+    DockerBase(String image, String suffix, String packageManager, String dockerfile) {
         this.image = image;
         this.suffix = suffix;
         this.packageManager = packageManager;
+        this.dockerfile = dockerfile;
     }
 
     public String getImage() {
@@ -57,4 +66,8 @@ public String getSuffix() {
     public String getPackageManager() {
         return packageManager;
     }
+
+    public String getDockerfile() {
+        return dockerfile;
+    }
 }