Merge branch 'main' of github.com:elastic/elasticsearch into custom-inference-service

Author: jonathan-buttner

.buildkite/scripts/third-party-test-credentials.sh

@@ -7,6 +7,7 @@ set -euo pipefail
 # The second/lowercase export is what the tests expect/require
 
 if [[ "${USE_3RD_PARTY_AZURE_CREDENTIALS:-}" == "true" ]]; then
+  # These credentials expire periodically and must be manually renewed - the process is in the onboarding/process docs.
   json=$(vault read -format=json secret/ci/elastic-elasticsearch/migrated/azure_thirdparty_test_creds)
 
   AZURE_STORAGE_ACCOUNT_SECRET=$(echo "$json" | jq -r .data.account_id)
@@ -19,6 +20,7 @@ if [[ "${USE_3RD_PARTY_AZURE_CREDENTIALS:-}" == "true" ]]; then
 fi
 
 if [[ "${USE_3RD_PARTY_AZURE_SAS_CREDENTIALS:-}" == "true" ]]; then
+  # These credentials expire periodically and must be manually renewed - the process is in the onboarding/process docs.
   json=$(vault read -format=json secret/ci/elastic-elasticsearch/migrated/azure_thirdparty_sas_test_creds)
 
   AZURE_STORAGE_ACCOUNT_SECRET=$(echo "$json" | jq -r .data.account_id)

benchmarks/src/main/java/org/elasticsearch/benchmark/compute/operator/ValuesAggregatorBenchmark.java

@@ -21,10 +21,13 @@
 import org.elasticsearch.compute.data.Block;
 import org.elasticsearch.compute.data.BlockFactory;
 import org.elasticsearch.compute.data.BytesRefBlock;
+import org.elasticsearch.compute.data.BytesRefVector;
 import org.elasticsearch.compute.data.ElementType;
 import org.elasticsearch.compute.data.IntBlock;
+import org.elasticsearch.compute.data.IntVector;
 import org.elasticsearch.compute.data.LongBlock;
 import org.elasticsearch.compute.data.LongVector;
+import org.elasticsearch.compute.data.OrdinalBytesRefVector;
 import org.elasticsearch.compute.data.Page;
 import org.elasticsearch.compute.operator.AggregationOperator;
 import org.elasticsearch.compute.operator.DriverContext;
@@ -282,11 +285,18 @@ private static Block dataBlock(int groups, String dataType) {
         int blockLength = blockLength(groups);
         return switch (dataType) {
             case BYTES_REF -> {
-                try (BytesRefBlock.Builder builder = blockFactory.newBytesRefBlockBuilder(blockLength)) {
+                try (
+                    BytesRefVector.Builder dict = blockFactory.newBytesRefVectorBuilder(blockLength);
+                    IntVector.Builder ords = blockFactory.newIntVectorBuilder(blockLength)
+                ) {
+                    final int dictLength = Math.min(blockLength, KEYWORDS.length);
+                    for (int i = 0; i < dictLength; i++) {
+                        dict.appendBytesRef(KEYWORDS[i]);
+                    }
                     for (int i = 0; i < blockLength; i++) {
-                        builder.appendBytesRef(KEYWORDS[i % KEYWORDS.length]);
+                        ords.appendInt(i % dictLength);
                     }
-                    yield builder.build();
+                    yield new OrdinalBytesRefVector(ords.build(), dict.build()).asBlock();
                 }
             }
             case INT -> {

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

@@ -17,7 +17,7 @@
 import org.apache.lucene.util.VectorUtil;
 import org.apache.lucene.util.quantization.OptimizedScalarQuantizer;
 import org.elasticsearch.common.logging.LogConfigurator;
-import org.elasticsearch.simdvec.internal.vectorization.ES91OSQVectorsScorer;
+import org.elasticsearch.simdvec.ES91OSQVectorsScorer;
 import org.elasticsearch.simdvec.internal.vectorization.ESVectorizationProvider;
 import org.openjdk.jmh.annotations.Benchmark;
 import org.openjdk.jmh.annotations.BenchmarkMode;

docs/changelog/127355.yaml

@@ -0,0 +1,5 @@
+pr: 127355
+summary: '`text ==` and `text !=` pushdown'
+area: ES|QL
+type: enhancement
+issues: []

docs/changelog/127629.yaml

@@ -0,0 +1,5 @@
+pr: 127629
+summary: ES|QL SAMPLE aggregation function
+area: Machine Learning
+type: feature
+issues: []

docs/changelog/127752.yaml

@@ -0,0 +1,6 @@
+pr: 127752
+summary: Downsampling does not consider passthrough fields as dimensions
+area: Downsampling
+type: bug
+issues:
+ - 125156

docs/changelog/127817.yaml

@@ -0,0 +1,5 @@
+pr: 127817
+summary: Replace auto-read with proper flow-control in HTTP pipeline
+area: Network
+type: enhancement
+issues: []

docs/changelog/127824.yaml

@@ -0,0 +1,6 @@
+pr: 127824
+summary: Skip the validation when retrieving the index mode during reindexing a time
+  series data stream
+area: TSDB
+type: bug
+issues: []

docs/changelog/127849.yaml

@@ -0,0 +1,5 @@
+pr: 127849
+summary: Optimize ordinal inputs in Values aggregation
+area: "ES|QL"
+type: enhancement
+issues: []

docs/changelog/127856.yaml

@@ -0,0 +1,5 @@
+pr: 127856
+summary: Fix services API Google Vertex AI Rerank location field requirement
+area: Machine Learning
+type: bug
+issues: []