Skip to content

Commit 2f2abfb

Browse files
prwhelanprwhelan
authored
Merge branch 'main' into sagemaker-elastic-update-fix
2 parents 189ce70 + bfab88c commit 2f2abfb

File tree

50 files changed

+1595
-1715
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

50 files changed

+1595
-1715
lines changed

‎benchmarks/src/main/java/org/elasticsearch/benchmark/_nightly/esql/ValuesSourceReaderBenchmark.java‎

Lines changed: 32 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -92,7 +92,8 @@ public class ValuesSourceReaderBenchmark {
9292
"double",
9393
"keyword",
9494
"stored_keyword",
95-
"3_stored_keywords" };
95+
"3_stored_keywords",
96+
"keyword_mv" };
9697

9798
private static final int BLOCK_LENGTH = 16 * 1024;
9899
private static final int INDEX_SIZE = 10 * BLOCK_LENGTH;
@@ -332,7 +333,7 @@ public FieldNamesFieldMapper.FieldNamesFieldType fieldNames() {
332333
@Param({ "in_order", "shuffled" })
333334
public String layout;
334335

335-
@Param({ "long", "keyword", "stored_keyword" })
336+
@Param({ "long", "keyword", "stored_keyword", "keyword_mv" })
336337
public String name;
337338

338339
private Directory directory;
@@ -398,6 +399,22 @@ public void benchmark() {
398399
}
399400
}
400401
}
402+
case "keyword_mv" -> {
403+
BytesRef scratch = new BytesRef();
404+
BytesRefBlock values = op.getOutput().<BytesRefBlock>getBlock(1);
405+
for (int p = 0; p < values.getPositionCount(); p++) {
406+
int count = values.getValueCount(p);
407+
if (count > 0) {
408+
int first = values.getFirstValueIndex(p);
409+
for (int i = 0; i < count; i++) {
410+
BytesRef r = values.getBytesRef(first + i, scratch);
411+
r.offset++;
412+
r.length--;
413+
sum += Integer.parseInt(r.utf8ToString());
414+
}
415+
}
416+
}
417+
}
401418
}
402419
}
403420
long expected = 0;
@@ -407,6 +424,16 @@ public void benchmark() {
407424
expected += i % 1000;
408425
}
409426
break;
427+
case "keyword_mv":
428+
for (int i = 0; i < INDEX_SIZE; i++) {
429+
int v1 = i % 1000;
430+
expected += v1;
431+
int v2 = i % 500;
432+
if (v1 != v2) {
433+
expected += v2;
434+
}
435+
}
436+
break;
410437
case "3_stored_keywords":
411438
for (int i = 0; i < INDEX_SIZE; i++) {
412439
expected += 3 * (i % 1000);
@@ -461,7 +488,9 @@ private void setupIndex() throws IOException {
461488
new StoredField("double", (double) i),
462489
new KeywordFieldMapper.KeywordField("keyword_1", new BytesRef(c + i % 1000), keywordFieldType),
463490
new KeywordFieldMapper.KeywordField("keyword_2", new BytesRef(c + i % 1000), keywordFieldType),
464-
new KeywordFieldMapper.KeywordField("keyword_3", new BytesRef(c + i % 1000), keywordFieldType)
491+
new KeywordFieldMapper.KeywordField("keyword_3", new BytesRef(c + i % 1000), keywordFieldType),
492+
new KeywordFieldMapper.KeywordField("keyword_mv", new BytesRef(c + i % 1000), keywordFieldType),
493+
new KeywordFieldMapper.KeywordField("keyword_mv", new BytesRef(c + i % 500), keywordFieldType)
465494
)
466495
);
467496
if (i % COMMIT_INTERVAL == 0) {

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

Lines changed: 5 additions & 21 deletions
Original file line numberDiff line numberDiff line change
@@ -95,8 +95,7 @@ static void selfTest() {
9595
try {
9696
for (String groups : ValuesAggregatorBenchmark.class.getField("groups").getAnnotationsByType(Param.class)[0].value()) {
9797
for (String dataType : ValuesAggregatorBenchmark.class.getField("dataType").getAnnotationsByType(Param.class)[0].value()) {
98-
run(Integer.parseInt(groups), dataType, 10, 0);
99-
run(Integer.parseInt(groups), dataType, 10, 1);
98+
run(Integer.parseInt(groups), dataType, 10);
10099
}
101100
}
102101
} catch (NoSuchFieldException e) {
@@ -114,10 +113,7 @@ static void selfTest() {
114113
@Param({ BYTES_REF, INT, LONG })
115114
public String dataType;
116115

117-
@Param({ "0", "1" })
118-
public int numOrdinalMerges;
119-
120-
private static Operator operator(DriverContext driverContext, int groups, String dataType, int numOrdinalMerges) {
116+
private static Operator operator(DriverContext driverContext, int groups, String dataType) {
121117
if (groups == 1) {
122118
return new AggregationOperator(
123119
List.of(supplier(dataType).aggregatorFactory(AggregatorMode.SINGLE, List.of(0)).apply(driverContext)),
@@ -132,20 +128,8 @@ private static Operator operator(DriverContext driverContext, int groups, String
132128
) {
133129
@Override
134130
public Page getOutput() {
135-
mergeOrdinal();
136131
return super.getOutput();
137132
}
138-
139-
// simulate OrdinalsGroupingOperator
140-
void mergeOrdinal() {
141-
var merged = supplier(dataType).groupingAggregatorFactory(AggregatorMode.SINGLE, List.of(1)).apply(driverContext);
142-
for (int i = 0; i < numOrdinalMerges; i++) {
143-
for (int p = 0; p < groups; p++) {
144-
merged.addIntermediateRow(p, aggregators.getFirst(), p);
145-
}
146-
}
147-
aggregators.set(0, merged);
148-
}
149133
};
150134
}
151135

@@ -352,12 +336,12 @@ private static Block groupingBlock(int groups) {
352336

353337
@Benchmark
354338
public void run() {
355-
run(groups, dataType, OP_COUNT, numOrdinalMerges);
339+
run(groups, dataType, OP_COUNT);
356340
}
357341

358-
private static void run(int groups, String dataType, int opCount, int numOrdinalMerges) {
342+
private static void run(int groups, String dataType, int opCount) {
359343
DriverContext driverContext = driverContext();
360-
try (Operator operator = operator(driverContext, groups, dataType, numOrdinalMerges)) {
344+
try (Operator operator = operator(driverContext, groups, dataType)) {
361345
Page page = page(groups, dataType);
362346
for (int i = 0; i < opCount; i++) {
363347
operator.addInput(page.shallowCopy());

‎docs/changelog/128639.yaml‎

Lines changed: 6 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,6 @@
1+
pr: 128639
2+
summary: Substitue `date_trunc` with `round_to` when the pre-calculated rounding points
3+
are available
4+
area: ES|QL
5+
type: enhancement
6+
issues: []

‎docs/changelog/131061.yaml‎

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,5 @@
1+
pr: 131061
2+
summary: Speed up reading multivalued keywords
3+
area: ES|QL
4+
type: enhancement
5+
issues: []

‎modules/ingest-geoip/src/main/java/org/elasticsearch/ingest/geoip/ConfigDatabases.java‎

Lines changed: 6 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -11,7 +11,7 @@
1111
import org.apache.logging.log4j.LogManager;
1212
import org.apache.logging.log4j.Logger;
1313
import org.elasticsearch.cluster.metadata.ProjectId;
14-
import org.elasticsearch.core.FixForMultiProject;
14+
import org.elasticsearch.core.NotMultiProjectCapable;
1515
import org.elasticsearch.env.Environment;
1616
import org.elasticsearch.watcher.FileChangesListener;
1717
import org.elasticsearch.watcher.FileWatcher;
@@ -34,6 +34,9 @@
3434
* Keeps track of user provided databases in the ES_HOME/config/ingest-geoip directory.
3535
* This directory is monitored and files updates are picked up and may cause databases being loaded or removed at runtime.
3636
*/
37+
@NotMultiProjectCapable(
38+
description = "Custom databases not available in serverless, we should review this class for MP again after serverless is enabled"
39+
)
3740
final class ConfigDatabases implements Closeable {
3841

3942
private static final Logger logger = LogManager.getLogger(ConfigDatabases.class);
@@ -71,7 +74,7 @@ Map<String, DatabaseReaderLazyLoader> getConfigDatabases() {
7174
return configDatabases;
7275
}
7376

74-
@FixForMultiProject(description = "Replace DEFAULT project")
77+
@NotMultiProjectCapable(description = "Replace DEFAULT project after serverless is enabled")
7578
void updateDatabase(Path file, boolean update) {
7679
String databaseFileName = file.getFileName().toString();
7780
try {
@@ -93,7 +96,7 @@ void updateDatabase(Path file, boolean update) {
9396
}
9497
}
9598

96-
@FixForMultiProject(description = "Replace DEFAULT project")
99+
@NotMultiProjectCapable(description = "Replace DEFAULT project after serverless is enabled")
97100
Map<String, DatabaseReaderLazyLoader> initConfigDatabases() throws IOException {
98101
Map<String, DatabaseReaderLazyLoader> databases = new HashMap<>();
99102

‎modules/ingest-geoip/src/main/java/org/elasticsearch/ingest/geoip/direct/TransportGetDatabaseConfigurationAction.java‎

Lines changed: 21 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -13,11 +13,14 @@
1313
import org.elasticsearch.action.FailedNodeException;
1414
import org.elasticsearch.action.support.ActionFilters;
1515
import org.elasticsearch.action.support.nodes.TransportNodesAction;
16+
import org.elasticsearch.cluster.metadata.ProjectMetadata;
1617
import org.elasticsearch.cluster.node.DiscoveryNode;
18+
import org.elasticsearch.cluster.project.ProjectResolver;
1719
import org.elasticsearch.cluster.service.ClusterService;
1820
import org.elasticsearch.common.io.stream.StreamInput;
1921
import org.elasticsearch.common.regex.Regex;
2022
import org.elasticsearch.ingest.geoip.DatabaseNodeService;
23+
import org.elasticsearch.ingest.geoip.GeoIpDownloaderTaskExecutor;
2124
import org.elasticsearch.ingest.geoip.GeoIpTaskState;
2225
import org.elasticsearch.ingest.geoip.IngestGeoIpMetadata;
2326
import org.elasticsearch.injection.guice.Inject;
@@ -48,14 +51,16 @@ public class TransportGetDatabaseConfigurationAction extends TransportNodesActio
4851
List<DatabaseConfigurationMetadata>> {
4952

5053
private final DatabaseNodeService databaseNodeService;
54+
private final ProjectResolver projectResolver;
5155

5256
@Inject
5357
public TransportGetDatabaseConfigurationAction(
5458
TransportService transportService,
5559
ClusterService clusterService,
5660
ThreadPool threadPool,
5761
ActionFilters actionFilters,
58-
DatabaseNodeService databaseNodeService
62+
DatabaseNodeService databaseNodeService,
63+
ProjectResolver projectResolver
5964
) {
6065
super(
6166
GetDatabaseConfigurationAction.NAME,
@@ -66,6 +71,7 @@ public TransportGetDatabaseConfigurationAction(
6671
threadPool.executor(ThreadPool.Names.MANAGEMENT)
6772
);
6873
this.databaseNodeService = databaseNodeService;
74+
this.projectResolver = projectResolver;
6975
}
7076

7177
protected List<DatabaseConfigurationMetadata> createActionContext(Task task, GetDatabaseConfigurationAction.Request request) {
@@ -82,25 +88,29 @@ protected List<DatabaseConfigurationMetadata> createActionContext(Task task, Get
8288
"wildcard only supports a single value, please use comma-separated values or a single wildcard value"
8389
);
8490
}
85-
8691
List<DatabaseConfigurationMetadata> results = new ArrayList<>();
87-
PersistentTasksCustomMetadata tasksMetadata = PersistentTasksCustomMetadata.getPersistentTasksCustomMetadata(
88-
clusterService.state()
89-
);
92+
ProjectMetadata projectMetadata = projectResolver.getProjectMetadata(clusterService.state());
93+
PersistentTasksCustomMetadata tasksMetadata = PersistentTasksCustomMetadata.get(projectMetadata);
94+
String geoIpTaskId = GeoIpDownloaderTaskExecutor.getTaskId(projectMetadata.id(), projectResolver.supportsMultipleProjects());
95+
9096
for (String id : ids) {
91-
results.addAll(getWebDatabases(tasksMetadata, id));
92-
results.addAll(getMaxmindDatabases(clusterService, id));
97+
results.addAll(getWebDatabases(geoIpTaskId, tasksMetadata, id));
98+
results.addAll(getMaxmindDatabases(projectMetadata, id));
9399
}
94100
return results;
95101
}
96102

97103
/*
98104
* This returns read-only database information about the databases managed by the standard downloader
99105
*/
100-
private static Collection<DatabaseConfigurationMetadata> getWebDatabases(PersistentTasksCustomMetadata tasksMetadata, String id) {
106+
private static Collection<DatabaseConfigurationMetadata> getWebDatabases(
107+
String geoIpTaskId,
108+
PersistentTasksCustomMetadata tasksMetadata,
109+
String id
110+
) {
101111
List<DatabaseConfigurationMetadata> webDatabases = new ArrayList<>();
102112
if (tasksMetadata != null) {
103-
PersistentTasksCustomMetadata.PersistentTask<?> maybeGeoIpTask = tasksMetadata.getTask("geoip-downloader");
113+
PersistentTasksCustomMetadata.PersistentTask<?> maybeGeoIpTask = tasksMetadata.getTask(geoIpTaskId);
104114
if (maybeGeoIpTask != null) {
105115
GeoIpTaskState geoIpTaskState = (GeoIpTaskState) maybeGeoIpTask.getState();
106116
if (geoIpTaskState != null) {
@@ -137,12 +147,9 @@ private static String getDatabaseNameForFileName(String databaseFileName) {
137147
/*
138148
* This returns information about databases that are downloaded from maxmind.
139149
*/
140-
private static Collection<DatabaseConfigurationMetadata> getMaxmindDatabases(ClusterService clusterService, String id) {
150+
private static Collection<DatabaseConfigurationMetadata> getMaxmindDatabases(ProjectMetadata projectMetadata, String id) {
141151
List<DatabaseConfigurationMetadata> maxmindDatabases = new ArrayList<>();
142-
final IngestGeoIpMetadata geoIpMeta = clusterService.state()
143-
.metadata()
144-
.getProject()
145-
.custom(IngestGeoIpMetadata.TYPE, IngestGeoIpMetadata.EMPTY);
152+
final IngestGeoIpMetadata geoIpMeta = projectMetadata.custom(IngestGeoIpMetadata.TYPE, IngestGeoIpMetadata.EMPTY);
146153
if (Regex.isSimpleMatchPattern(id)) {
147154
for (Map.Entry<String, DatabaseConfigurationMetadata> entry : geoIpMeta.getDatabases().entrySet()) {
148155
if (Regex.simpleMatch(id, entry.getKey())) {

‎muted-tests.yml‎

Lines changed: 0 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -415,9 +415,6 @@ tests:
415415
- class: org.elasticsearch.xpack.esql.action.EsqlActionBreakerIT
416416
method: testRowStatsProjectGroupByInt
417417
issue: https://github.com/elastic/elasticsearch/issues/131024
418-
- class: org.elasticsearch.xpack.esql.action.CrossClusterAsyncQueryStopIT
419-
method: testStopQueryLocal
420-
issue: https://github.com/elastic/elasticsearch/issues/121672
421418
- class: org.elasticsearch.xpack.esql.qa.multi_node.EsqlSpecIT
422419
method: test {lookup-join.MvJoinKeyOnFromAfterStats ASYNC}
423420
issue: https://github.com/elastic/elasticsearch/issues/131148

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

Lines changed: 39 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -656,7 +656,7 @@ public BlockLoader.Block read(BlockFactory factory, Docs docs) throws IOExceptio
656656
if (docs.count() == 1) {
657657
return readSingleDoc(factory, docs.get(0));
658658
}
659-
try (BlockLoader.SingletonOrdinalsBuilder builder = factory.singletonOrdinalsBuilder(ordinals, docs.count())) {
659+
try (var builder = factory.singletonOrdinalsBuilder(ordinals, docs.count())) {
660660
for (int i = 0; i < docs.count(); i++) {
661661
int doc = docs.get(i);
662662
if (doc < ordinals.docID()) {
@@ -701,13 +701,29 @@ private static class Ordinals extends BlockDocValuesReader {
701701

702702
@Override
703703
public BlockLoader.Block read(BlockFactory factory, Docs docs) throws IOException {
704-
try (BytesRefBuilder builder = factory.bytesRefsFromDocValues(docs.count())) {
704+
if (docs.count() == 1) {
705+
return readSingleDoc(factory, docs.get(0));
706+
}
707+
try (var builder = factory.sortedSetOrdinalsBuilder(ordinals, docs.count())) {
705708
for (int i = 0; i < docs.count(); i++) {
706709
int doc = docs.get(i);
707710
if (doc < ordinals.docID()) {
708711
throw new IllegalStateException("docs within same block must be in order");
709712
}
710-
read(doc, builder);
713+
if (ordinals.advanceExact(doc) == false) {
714+
builder.appendNull();
715+
continue;
716+
}
717+
int count = ordinals.docValueCount();
718+
if (count == 1) {
719+
builder.appendOrd(Math.toIntExact(ordinals.nextOrd()));
720+
} else {
721+
builder.beginPositionEntry();
722+
for (int c = 0; c < count; c++) {
723+
builder.appendOrd(Math.toIntExact(ordinals.nextOrd()));
724+
}
725+
builder.endPositionEntry();
726+
}
711727
}
712728
return builder.build();
713729
}
@@ -718,6 +734,26 @@ public void read(int docId, BlockLoader.StoredFields storedFields, Builder build
718734
read(docId, (BytesRefBuilder) builder);
719735
}
720736

737+
private BlockLoader.Block readSingleDoc(BlockFactory factory, int docId) throws IOException {
738+
if (ordinals.advanceExact(docId) == false) {
739+
return factory.constantNulls();
740+
}
741+
int count = ordinals.docValueCount();
742+
if (count == 1) {
743+
BytesRef v = ordinals.lookupOrd(ordinals.nextOrd());
744+
return factory.constantBytes(BytesRef.deepCopyOf(v));
745+
}
746+
try (var builder = factory.bytesRefsFromDocValues(count)) {
747+
builder.beginPositionEntry();
748+
for (int c = 0; c < count; c++) {
749+
BytesRef v = ordinals.lookupOrd(ordinals.nextOrd());
750+
builder.appendBytesRef(v);
751+
}
752+
builder.endPositionEntry();
753+
return builder.build();
754+
}
755+
}
756+
721757
private void read(int docId, BytesRefBuilder builder) throws IOException {
722758
if (false == ordinals.advanceExact(docId)) {
723759
builder.appendNull();

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

Lines changed: 12 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -417,11 +417,14 @@ interface BlockFactory {
417417
Block constantBytes(BytesRef value);
418418

419419
/**
420-
* Build a reader for reading keyword ordinals.
420+
* Build a reader for reading {@link SortedDocValues}
421421
*/
422422
SingletonOrdinalsBuilder singletonOrdinalsBuilder(SortedDocValues ordinals, int count);
423423

424-
// TODO support non-singleton ords
424+
/**
425+
* Build a reader for reading {@link SortedSetDocValues}
426+
*/
427+
SortedSetOrdinalsBuilder sortedSetOrdinalsBuilder(SortedSetDocValues ordinals, int count);
425428

426429
AggregateMetricDoubleBuilder aggregateMetricDoubleBuilder(int count);
427430
}
@@ -509,6 +512,13 @@ interface SingletonOrdinalsBuilder extends Builder {
509512
SingletonOrdinalsBuilder appendOrd(int value);
510513
}
511514

515+
interface SortedSetOrdinalsBuilder extends Builder {
516+
/**
517+
* Appends an ordinal to the builder.
518+
*/
519+
SortedSetOrdinalsBuilder appendOrd(int value);
520+
}
521+
512522
interface AggregateMetricDoubleBuilder extends Builder {
513523

514524
DoubleBuilder min();

0 commit comments

Comments
 (0)