Merge branch 'main' into expressionJoin_v4

Author: julian-elastic

.buildkite/pipelines/java-ea-check-new-build.yml

@@ -0,0 +1,7 @@
+steps:
+  - command: .buildkite/scripts/trigger-if-java-ea-new-build.sh
+    env:
+      RECENT_TIME_WINDOW: "24" # time window to consider a build as new in hours
+    agents:
+      image: "docker.elastic.co/ci-agent-images/eck-region/buildkite-agent:1.5"
+      memory: "4G"

.buildkite/pipelines/periodic-fwc.template.yml

@@ -1,5 +1,5 @@
 steps:
-  - label: "{{matrix.FWC_VERSION}}" / fwc
+  - label: "{{matrix.FWC_VERSION}} / fwc"
     command: .ci/scripts/run-gradle.sh -Dbwc.checkout.align=true v$$FWC_VERSION#fwcTest -Dtests.bwc.snapshot=false
     timeout_in_minutes: 300
     agents:

.buildkite/pipelines/periodic-fwc.yml

@@ -1,6 +1,6 @@
 # This file is auto-generated. See .buildkite/pipelines/periodic-fwc.template.yml
 steps:
-  - label: "{{matrix.FWC_VERSION}}" / fwc
+  - label: "{{matrix.FWC_VERSION}} / fwc"
     command: .ci/scripts/run-gradle.sh -Dbwc.checkout.align=true v$$FWC_VERSION#fwcTest -Dtests.bwc.snapshot=false
     timeout_in_minutes: 300
     agents:

.buildkite/pipelines/pull-request-transport-versions.yml

@@ -0,0 +1 @@
+steps: []

.buildkite/scripts/trigger-if-java-ea-new-build.sh

@@ -0,0 +1,95 @@
+#
+# 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".
+#
+
+#!/bin/bash
+
+# Allow overriding the time window (in hours) to check for new builds, defaults to 24
+RECENT_TIME_WINDOW=${RECENT_TIME_WINDOW:-24}
+
+# Extract current JDK major version from bundled_jdk in version.properties
+CURRENT_JDK=$(grep "^bundled_jdk =" build-tools-internal/version.properties | cut -d'=' -f2 | tr -d ' ' | cut -d'.' -f1)
+TARGET_JDK=$((CURRENT_JDK + 1))
+
+echo "Current JDK major version: $CURRENT_JDK"
+echo "Target JDK major version: $TARGET_JDK"
+
+# Query Elasticsearch JDK archive for available JDKs
+JDK_ARCHIVE_URL="https://builds.es-jdk-archive.com/jdks/openjdk/recent.json"
+echo "Querying JDK archive: $JDK_ARCHIVE_URL"
+
+# Fetch JDK info and filter for target major version
+JDK_DATA=$(curl -s "$JDK_ARCHIVE_URL")
+
+if [[ -z "$JDK_DATA" ]]; then
+  echo "Failed to fetch JDK data from archive"
+  exit 1
+fi
+
+# Find the latest build for the target JDK version
+LATEST_BUILD=$(echo "$JDK_DATA" | jq -r --arg target "$TARGET_JDK" '
+  .majors[$target].builds |
+  sort_by(.archived_at) |
+  last'
+)
+
+if [[ "$LATEST_BUILD" == "null" || -z "$LATEST_BUILD" ]]; then
+  echo "No builds found for JDK $TARGET_JDK"
+  exit 1
+fi
+
+# Extract timestamp and JDK identifier
+TIMESTAMP=$(echo "$LATEST_BUILD" | jq -r '.archived_at')
+JDK_IDENTIFIER=$(echo "$LATEST_BUILD" | jq -r '.id')
+
+echo "Latest JDK ${TARGET_JDK} build from ES archive:"
+echo "  Timestamp: $TIMESTAMP"
+echo "  JDK Identifier: $JDK_IDENTIFIER"
+
+# Set variables for use in the pipeline trigger
+jdkbuild="$JDK_IDENTIFIER"
+jdk_timestamp="$TIMESTAMP"
+
+# Check if timestamp is within last 24 hours
+CURRENT_TIME=$(date +%s)
+BUILD_TIME=$(date -d "$TIMESTAMP" +%s 2>/dev/null || date -j -f "%Y-%m-%dT%H:%M:%S" "${TIMESTAMP%Z}" +%s 2>/dev/null || echo "0")
+
+if [[ "$BUILD_TIME" == "0" ]]; then
+  echo "Failed to parse timestamp: $TIMESTAMP"
+  SHOULD_TRIGGER="false"
+else
+  TIME_DIFF=$((CURRENT_TIME - BUILD_TIME))
+  TIME_WINDOW=$((RECENT_TIME_WINDOW * 60 * 60))
+
+  if [[ $TIME_DIFF -lt $TIME_WINDOW ]]; then
+    echo "Build is recent (less than ${RECENT_TIME_WINDOW}h old)"
+    SHOULD_TRIGGER="true"
+  else
+    echo "Build is older than ${RECENT_TIME_WINDOW} hours"
+    SHOULD_TRIGGER="false"
+  fi
+fi
+
+echo "SHOULD_TRIGGER: $SHOULD_TRIGGER"
+
+
+if [[ "$SHOULD_TRIGGER" == "true" ]]; then
+  EFFECTIVE_START_DATE=$(date -u -d "@$BUILD_TIME" +"%Y-%m-%dT%H:%M:%SZ" 2>/dev/null || date -u -r "$BUILD_TIME" +"%Y-%m-%dT%H:%M:%SZ" 2>/dev/null || echo "")
+  echo "Triggering performance-esbench-jdk for new jdk build $JDK_IDENTIFIER"
+  cat << EOF | buildkite-agent pipeline upload
+steps:
+- trigger: elasticsearch-performance-esbench-jdk
+  label: Triggering performance-esbench-jdk for new jdk build $JDK_IDENTIFIER
+  async: true
+  build:
+    branch: "$BUILDKITE_BRANCH"
+    env:
+      EFFECTIVE_START_DATE: "$EFFECTIVE_START_DATE"
+      EXECUTION_MODE: "start-run"
+EOF
+fi

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

@@ -0,0 +1,77 @@
+/*
+ * 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.vector;
+
+import org.elasticsearch.common.logging.LogConfigurator;
+import org.elasticsearch.index.codec.vectors.cluster.NeighborHood;
+import org.openjdk.jmh.annotations.Benchmark;
+import org.openjdk.jmh.annotations.BenchmarkMode;
+import org.openjdk.jmh.annotations.Fork;
+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.Warmup;
+import org.openjdk.jmh.infra.Blackhole;
+
+import java.io.IOException;
+import java.util.Random;
+import java.util.concurrent.TimeUnit;
+
+@BenchmarkMode(Mode.AverageTime)
+@OutputTimeUnit(TimeUnit.SECONDS)
+@State(Scope.Benchmark)
+// first iteration is complete garbage, so make sure we really warmup
+@Warmup(iterations = 1, time = 1)
+// real iterations. not useful to spend tons of time here, better to fork more
+@Measurement(iterations = 3, time = 1)
+// engage some noise reduction
+@Fork(value = 1)
+public class ComputeNeighboursBenchmark {
+
+    static {
+        LogConfigurator.configureESLogging(); // native access requires logging to be initialized
+    }
+
+    @Param({ "1000", "2000", "3000", "5000", "10000", "20000", "50000" })
+    int numVectors;
+
+    @Param({ "384", "782", "1024" })
+    int dims;
+
+    float[][] vectors;
+    int clusterPerNeighbour = 128;
+
+    @Setup
+    public void setup() throws IOException {
+        Random random = new Random(123);
+        vectors = new float[numVectors][dims];
+        for (float[] vector : vectors) {
+            for (int i = 0; i < dims; i++) {
+                vector[i] = random.nextFloat();
+            }
+        }
+    }
+
+    @Benchmark
+    @Fork(jvmArgsPrepend = { "--add-modules=jdk.incubator.vector" })
+    public void bruteForce(Blackhole bh) {
+        bh.consume(NeighborHood.computeNeighborhoodsBruteForce(vectors, clusterPerNeighbour));
+    }
+
+    @Benchmark
+    @Fork(jvmArgsPrepend = { "--add-modules=jdk.incubator.vector" })
+    public void graph(Blackhole bh) throws IOException {
+        bh.consume(NeighborHood.computeNeighborhoodsGraph(vectors, clusterPerNeighbour));
+    }
+}

build-tools-internal/src/integTest/groovy/org/elasticsearch/gradle/internal/ElasticsearchTestBasePluginFuncTest.groovy

@@ -108,4 +108,65 @@ class ElasticsearchTestBasePluginFuncTest extends AbstractGradleFuncTest {
         then:
         result.task(':test').outcome == TaskOutcome.UP_TO_DATE
     }
+
+    def "uses new test seed for every invocation"() {
+        given:
+        file("src/test/java/acme/SomeTests.java").text = """
+
+        public class SomeTests {
+            @org.junit.Test
+            public void printTestSeed() {
+                System.out.println("TESTSEED=[" + System.getProperty("tests.seed") + "]");
+            }
+        }
+
+        """
+        buildFile.text = """
+            plugins {
+             id 'java'
+             id 'elasticsearch.test-base'
+            }
+
+            tasks.named('test').configure {
+                testLogging {
+                    showStandardStreams = true
+                }
+            }
+
+            tasks.register('test2', Test) {
+                classpath = sourceSets.test.runtimeClasspath
+                testClassesDirs = sourceSets.test.output.classesDirs
+                testLogging {
+                    showStandardStreams = true
+                }
+            }
+
+            repositories {
+                mavenCentral()
+            }
+
+            dependencies {
+                testImplementation 'junit:junit:4.12'
+            }
+
+        """
+
+        when:
+        def result1 = gradleRunner("cleanTest", "cleanTest2", "test", "test2").build()
+        def result2 = gradleRunner("cleanTest", "cleanTest2", "test", "test2").build()
+
+        then:
+        def seeds1 = result1.output.findAll(/(?m)TESTSEED=\[([^\]]+)\]/) { it[1] }
+        def seeds2 = result2.output.findAll(/(?m)TESTSEED=\[([^\]]+)\]/) { it[1] }
+
+        seeds1.unique().size() == 1
+        seeds2.unique().size() == 1
+
+        verifyAll {
+            seeds1[0] != null
+            seeds2[0] != null
+            seeds1[0] != seeds2[0]
+        }
+        result2.output.contains("Configuration cache entry reused.")
+    }
 }

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

@@ -24,6 +24,7 @@
 import org.gradle.api.Project;
 import org.gradle.api.Task;
 import org.gradle.api.artifacts.Configuration;
+import org.gradle.api.configuration.BuildFeatures;
 import org.gradle.api.file.FileCollection;
 import org.gradle.api.plugins.JavaPlugin;
 import org.gradle.api.provider.ProviderFactory;
@@ -56,6 +57,9 @@ public abstract class ElasticsearchTestBasePlugin implements Plugin<Project> {
     @Inject
     protected abstract ProviderFactory getProviderFactory();
 
+    @Inject
+    protected abstract BuildFeatures getBuildFeatures();
+
     @Override
     public void apply(Project project) {
         project.getRootProject().getPlugins().apply(GlobalBuildInfoPlugin.class);
@@ -164,9 +168,11 @@ public void execute(Task t) {
             );
             test.systemProperties(sysprops);
 
-            // ignore changing test seed when build is passed -Dignore.tests.seed for cacheability experimentation
-            if (System.getProperty("ignore.tests.seed") != null) {
-                nonInputProperties.systemProperty("tests.seed", buildParams.get().getTestSeed());
+            // ignore changing test seed when build is passed -Dignore.tests.seed for cacheability
+            // also ignore when configuration cache is on since the test seed as task input would break
+            // configuration cache reuse.
+            if (System.getProperty("ignore.tests.seed") != null || getBuildFeatures().getConfigurationCache().getActive().get()) {
+                nonInputProperties.systemProperty("tests.seed", buildParams.get().getTestSeedProvider());
             } else {
                 test.systemProperty("tests.seed", buildParams.get().getTestSeed());
             }

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

@@ -56,6 +56,8 @@ public interface BuildParameterExtension {
 
     String getTestSeed();
 
+    Provider<String> getTestSeedProvider();
+
     Boolean getCi();
 
     Integer getDefaultParallel();
@@ -66,7 +68,7 @@ public interface BuildParameterExtension {
 
     Provider<BwcVersions> getBwcVersionsProvider();
 
-    Random getRandom();
+    Provider<Random> getRandom();
 
     Boolean getGraalVmRuntime();
 }

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

@@ -39,7 +39,7 @@ public abstract class DefaultBuildParameterExtension implements BuildParameterEx
     private final Provider<String> gitRevision;
 
     private transient AtomicReference<ZonedDateTime> buildDate = new AtomicReference<>();
-    private final String testSeed;
+    private final Provider<String> testSeed;
     private final Boolean isCi;
     private final Integer defaultParallel;
     private final Boolean snapshotBuild;
@@ -58,7 +58,7 @@ public DefaultBuildParameterExtension(
         JavaVersion gradleJavaVersion,
         Provider<String> gitRevision,
         Provider<String> gitOrigin,
-        String testSeed,
+        Provider<String> testSeed,
         boolean isCi,
         int defaultParallel,
         final boolean isSnapshotBuild,
@@ -181,6 +181,11 @@ public ZonedDateTime getBuildDate() {
 
     @Override
     public String getTestSeed() {
+        return testSeed.get();
+    }
+
+    @Override
+    public Provider<String> getTestSeedProvider() {
         return testSeed;
     }
 
@@ -205,8 +210,8 @@ public BwcVersions getBwcVersions() {
     }
 
     @Override
-    public Random getRandom() {
-        return new Random(Long.parseUnsignedLong(testSeed.split(":")[0], 16));
+    public Provider<Random> getRandom() {
+        return getTestSeedProvider().map(seed -> new Random(Long.parseUnsignedLong(seed.split(":")[0], 16)));
     }
 
     @Override