feat(search): update to support OpenSearch 2.x

datahub-project · Sep 19, 2023 · 5f5147f · 5f5147f
1 parent e75900b
commit 5f5147f
Show file tree

Hide file tree

Showing 182 changed files with 2,913 additions and 1,584 deletions.
diff --git a/build.gradle b/build.gradle
@@ -8,7 +8,7 @@ buildscript {
   ext.openTelemetryVersion = '1.18.0'
   ext.neo4jVersion = '4.4.9'
   ext.testContainersVersion = '1.17.4'
-  ext.elasticsearchVersion = '7.10.2'
+  ext.elasticsearchVersion = '2.9.0' // ES 7.10, Opensearch 1.x, 2.x
   ext.jacksonVersion = '2.15.2'
   ext.jettyVersion = '9.4.46.v20220331'
   ext.playVersion = '2.8.18'
@@ -90,8 +90,8 @@ project.ext.externalDependency = [
     'ebean': 'io.ebean:ebean:' + ebeanVersion,
     'ebeanAgent': 'io.ebean:ebean-agent:' + ebeanVersion,
     'ebeanDdl': 'io.ebean:ebean-ddl-generator:' + ebeanVersion,
-    'elasticSearchRest': 'org.elasticsearch.client:elasticsearch-rest-high-level-client:' + elasticsearchVersion,
-    'elasticSearchTransport': 'org.elasticsearch.client:transport:' + elasticsearchVersion,
+    'elasticSearchRest': 'org.opensearch.client:opensearch-rest-high-level-client:' + elasticsearchVersion,
+    'elasticSearchJava': 'org.opensearch.client:opensearch-java:2.6.0',
     'findbugsAnnotations': 'com.google.code.findbugs:annotations:3.0.1',
     'graphqlJava': 'com.graphql-java:graphql-java:19.5',
     'graphqlJavaScalars': 'com.graphql-java:graphql-java-extended-scalars:19.1',
@@ -202,13 +202,15 @@ project.ext.externalDependency = [
     'springActuator': "org.springframework.boot:spring-boot-starter-actuator:$springBootVersion",
     'swaggerAnnotations': 'io.swagger.core.v3:swagger-annotations:2.1.12',
     'swaggerCli': 'io.swagger.codegen.v3:swagger-codegen-cli:3.0.41',
-    'testng': 'org.testng:testng:7.3.0',
+    'testngJava8': 'org.testng:testng:7.5.1',
+    'testng': 'org.testng:testng:7.8.0',
     'testContainers': 'org.testcontainers:testcontainers:' + testContainersVersion,
     'testContainersJunit': 'org.testcontainers:junit-jupiter:' + testContainersVersion,
     'testContainersPostgresql':'org.testcontainers:postgresql:' + testContainersVersion,
     'testContainersElasticsearch': 'org.testcontainers:elasticsearch:' + testContainersVersion,
     'testContainersCassandra': 'org.testcontainers:cassandra:' + testContainersVersion,
     'testContainersKafka': 'org.testcontainers:kafka:' + testContainersVersion,
+    'testContainersOpenSearch': 'org.opensearch:opensearch-testcontainers:2.0.0',
     'typesafeConfig':'com.typesafe:config:1.4.1',
     'wiremock':'com.github.tomakehurst:wiremock:2.10.0',
     'zookeeper': 'org.apache.zookeeper:zookeeper:3.4.14',
@@ -257,7 +259,6 @@ subprojects {
 
   plugins.withType(JavaPlugin) {
     dependencies {
-      testImplementation externalDependency.testng
       constraints {
         implementation('io.netty:netty-all:4.1.86.Final')
         implementation('org.apache.commons:commons-compress:1.21')
@@ -268,12 +269,6 @@ subprojects {
       }
     }
 
-    tasks.withType(Test) {
-      if (!name.startsWith('integ')) {
-        useTestNG()
-      }
-    }
-
     checkstyle {
       configDirectory = file("${project.rootDir}/gradle/checkstyle")
       sourceSets = [ getProject().sourceSets.main, getProject().sourceSets.test ]
@@ -292,6 +287,8 @@ subprojects {
     javaLauncher = javaToolchains.launcherFor {
       languageVersion = JavaLanguageVersion.of(11)
     }
+    // https://docs.gradle.org/current/userguide/performance.html
+    maxParallelForks = Runtime.runtime.availableProcessors().intdiv(2) ?: 1
   }
 
   afterEvaluate {

diff --git a/datahub-graphql-core/build.gradle b/datahub-graphql-core/build.gradle
@@ -24,6 +24,7 @@ dependencies {
     annotationProcessor externalDependency.lombok
 
     testImplementation externalDependency.mockito
+    testImplementation externalDependency.testng
 }
 
 graphqlCodegen {

diff --git a/...l-core/src/main/java/com/linkedin/datahub/graphql/analytics/service/AnalyticsService.java b/...l-core/src/main/java/com/linkedin/datahub/graphql/analytics/service/AnalyticsService.java
@@ -20,25 +20,25 @@
 import javax.annotation.Nonnull;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
-import org.elasticsearch.action.search.SearchRequest;
-import org.elasticsearch.action.search.SearchResponse;
-import org.elasticsearch.client.RequestOptions;
-import org.elasticsearch.client.RestHighLevelClient;
-import org.elasticsearch.index.query.BoolQueryBuilder;
-import org.elasticsearch.index.query.QueryBuilder;
-import org.elasticsearch.index.query.QueryBuilders;
-import org.elasticsearch.search.aggregations.AggregationBuilder;
-import org.elasticsearch.search.aggregations.AggregationBuilders;
-import org.elasticsearch.search.aggregations.Aggregations;
-import org.elasticsearch.search.aggregations.BucketOrder;
-import org.elasticsearch.search.aggregations.bucket.MultiBucketsAggregation;
-import org.elasticsearch.search.aggregations.bucket.filter.Filter;
-import org.elasticsearch.search.aggregations.bucket.histogram.DateHistogramInterval;
-import org.elasticsearch.search.aggregations.bucket.histogram.Histogram;
-import org.elasticsearch.search.aggregations.bucket.terms.Terms;
-import org.elasticsearch.search.aggregations.bucket.terms.TermsAggregationBuilder;
-import org.elasticsearch.search.aggregations.metrics.Cardinality;
-import org.elasticsearch.search.builder.SearchSourceBuilder;
+import org.opensearch.action.search.SearchRequest;
+import org.opensearch.action.search.SearchResponse;
+import org.opensearch.client.RequestOptions;
+import org.opensearch.client.RestHighLevelClient;
+import org.opensearch.index.query.BoolQueryBuilder;
+import org.opensearch.index.query.QueryBuilder;
+import org.opensearch.index.query.QueryBuilders;
+import org.opensearch.search.aggregations.AggregationBuilder;
+import org.opensearch.search.aggregations.AggregationBuilders;
+import org.opensearch.search.aggregations.Aggregations;
+import org.opensearch.search.aggregations.BucketOrder;
+import org.opensearch.search.aggregations.bucket.MultiBucketsAggregation;
+import org.opensearch.search.aggregations.bucket.filter.Filter;
+import org.opensearch.search.aggregations.bucket.histogram.DateHistogramInterval;
+import org.opensearch.search.aggregations.bucket.histogram.Histogram;
+import org.opensearch.search.aggregations.bucket.terms.Terms;
+import org.opensearch.search.aggregations.bucket.terms.TermsAggregationBuilder;
+import org.opensearch.search.aggregations.metrics.Cardinality;
+import org.opensearch.search.builder.SearchSourceBuilder;
 
 
 @Slf4j

diff --git a/...-graphql-core/src/test/java/com/linkedin/datahub/graphql/resolvers/ResolverUtilsTest.java b/...-graphql-core/src/test/java/com/linkedin/datahub/graphql/resolvers/ResolverUtilsTest.java
@@ -15,7 +15,6 @@
 import com.linkedin.metadata.query.filter.CriterionArray;
 import com.linkedin.metadata.query.filter.Filter;
 import graphql.schema.DataFetchingEnvironment;
-import junit.framework.TestCase;
 import org.testng.annotations.Test;
 import org.mockito.Mockito;
 
@@ -24,9 +23,10 @@
 import java.util.stream.Collectors;
 
 import static com.linkedin.datahub.graphql.resolvers.ResolverUtils.*;
+import static org.testng.AssertJUnit.assertEquals;
 
 
-public class ResolverUtilsTest extends TestCase {
+public class ResolverUtilsTest {
 
   @Test
   public void testCriterionFromFilter() throws Exception {

diff --git a/...c/test/java/com/linkedin/datahub/graphql/resolvers/auth/ListAccessTokensResolverTest.java b/...c/test/java/com/linkedin/datahub/graphql/resolvers/auth/ListAccessTokensResolverTest.java
@@ -10,15 +10,15 @@
 import com.linkedin.metadata.Constants;
 import graphql.schema.DataFetchingEnvironment;
 import java.util.Collections;
-import junit.framework.TestCase;
 import org.mockito.Mockito;
+import org.testng.annotations.Test;
 
 import static com.linkedin.datahub.graphql.resolvers.ResolverUtils.*;
 
 
-public class ListAccessTokensResolverTest extends TestCase {
+public class ListAccessTokensResolverTest {
 
-//  @Test
+  @Test
   public void testGetSuccess() throws Exception {
     final DataFetchingEnvironment mockEnv = Mockito.mock(DataFetchingEnvironment.class);
     final QueryContext mockAllowContext = TestUtils.getMockAllowContext();
@@ -36,13 +36,13 @@ public void testGetSuccess() throws Exception {
     Mockito.when(mockEnv.getArgument(Mockito.eq("input"))).thenReturn(input);
 
     final EntityClient mockClient = Mockito.mock(EntityClient.class);
-    Mockito.when(mockClient.filter(
+    Mockito.when(Mockito.eq(mockClient.filter(
         Mockito.eq(Constants.ACCESS_TOKEN_ENTITY_NAME),
             Mockito.eq(buildFilter(filters, Collections.emptyList())),
             Mockito.notNull(),
             Mockito.eq(input.getStart()),
             Mockito.eq(input.getCount()),
-            Mockito.eq(getAuthentication(mockEnv))))
+            Mockito.eq(getAuthentication(mockEnv)))))
         .thenReturn(null);
 
     final ListAccessTokensResolver resolver = new ListAccessTokensResolver(mockClient);

diff --git a/.../com/linkedin/datahub/graphql/resolvers/ingest/source/GetIngestionSourceResolverTest.java b/.../com/linkedin/datahub/graphql/resolvers/ingest/source/GetIngestionSourceResolverTest.java
@@ -14,11 +14,12 @@
 import com.linkedin.r2.RemoteInvocationException;
 import graphql.schema.DataFetchingEnvironment;
 import java.util.HashSet;
+
 import org.mockito.Mockito;
 import org.testng.annotations.Test;
 
-import static org.testng.Assert.*;
 import static com.linkedin.datahub.graphql.resolvers.ingest.IngestTestUtils.*;
+import static org.testng.Assert.assertThrows;
 
 public class GetIngestionSourceResolverTest {
 

diff --git a/...l-core/src/test/java/com/linkedin/datahub/graphql/resolvers/mutate/SiblingsUtilsTest.java b/...l-core/src/test/java/com/linkedin/datahub/graphql/resolvers/mutate/SiblingsUtilsTest.java
@@ -6,16 +6,16 @@
 import com.linkedin.common.urn.UrnUtils;
 import com.linkedin.datahub.graphql.resolvers.mutate.util.SiblingsUtils;
 import com.linkedin.metadata.entity.EntityService;
-import junit.framework.TestCase;
 import org.mockito.Mockito;
 import org.testng.annotations.Test;
 
 import java.util.HashSet;
 import java.util.Optional;
 
 import static com.linkedin.metadata.Constants.SIBLINGS_ASPECT_NAME;
+import static org.testng.AssertJUnit.assertEquals;
 
-public class SiblingsUtilsTest extends TestCase {
+public class SiblingsUtilsTest {
 
   private static final String TEST_DATASET_URN1 = "urn:li:dataset:(urn:li:dataPlatform:hive,fct_cypress_users_created,PROD)";
   private static final String TEST_DATASET_URN2 = "urn:li:dataset:(urn:li:dataPlatform:hive,fct_cypress_users_created2,PROD)";

diff --git a/datahub-graphql-core/src/test/java/com/linkedin/datahub/graphql/utils/DateUtilTest.java b/datahub-graphql-core/src/test/java/com/linkedin/datahub/graphql/utils/DateUtilTest.java
@@ -1,12 +1,13 @@
 package com.linkedin.datahub.graphql.utils;
 
 import com.linkedin.datahub.graphql.util.DateUtil;
-import junit.framework.TestCase;
 import org.joda.time.DateTime;
 import org.mockito.Mockito;
 import org.testng.annotations.Test;
 
-public class DateUtilTest extends TestCase {
+import static org.testng.AssertJUnit.assertEquals;
+
+public class DateUtilTest {
 
     private DateTime setTimeParts(int dayOfMonth, boolean zeroTime) {
         DateTime result = new DateTime()

diff --git a/datahub-upgrade/src/main/java/com/linkedin/datahub/upgrade/config/NoCodeCleanupConfig.java b/datahub-upgrade/src/main/java/com/linkedin/datahub/upgrade/config/NoCodeCleanupConfig.java
@@ -5,7 +5,7 @@
 import com.linkedin.metadata.utils.elasticsearch.IndexConvention;
 import io.ebean.Database;
 import javax.annotation.Nonnull;
-import org.elasticsearch.client.RestHighLevelClient;
+import org.opensearch.client.RestHighLevelClient;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.context.ApplicationContext;
 import org.springframework.context.annotation.Bean;

diff --git a/...c/main/java/com/linkedin/datahub/upgrade/nocodecleanup/DeleteLegacySearchIndicesStep.java b/...c/main/java/com/linkedin/datahub/upgrade/nocodecleanup/DeleteLegacySearchIndicesStep.java
@@ -7,9 +7,9 @@
 import com.linkedin.metadata.utils.elasticsearch.IndexConvention;
 import java.util.function.Function;
 import lombok.RequiredArgsConstructor;
-import org.elasticsearch.action.admin.indices.delete.DeleteIndexRequest;
-import org.elasticsearch.client.RequestOptions;
-import org.elasticsearch.client.RestHighLevelClient;
+import org.opensearch.action.admin.indices.delete.DeleteIndexRequest;
+import org.opensearch.client.RequestOptions;
+import org.opensearch.client.RestHighLevelClient;
 
 
 // Do we need SQL-tech specific migration paths?

diff --git a/...pgrade/src/main/java/com/linkedin/datahub/upgrade/nocodecleanup/NoCodeCleanupUpgrade.java b/...pgrade/src/main/java/com/linkedin/datahub/upgrade/nocodecleanup/NoCodeCleanupUpgrade.java
@@ -9,7 +9,7 @@
 import java.util.ArrayList;
 import java.util.Collections;
 import java.util.List;
-import org.elasticsearch.client.RestHighLevelClient;
+import org.opensearch.client.RestHighLevelClient;
 
 
 public class NoCodeCleanupUpgrade implements Upgrade {

diff --git a/...in/java/com/linkedin/datahub/upgrade/system/elasticsearch/steps/BuildIndicesPostStep.java b/...in/java/com/linkedin/datahub/upgrade/system/elasticsearch/steps/BuildIndicesPostStep.java
@@ -16,8 +16,8 @@
 
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
-import org.elasticsearch.action.admin.indices.settings.put.UpdateSettingsRequest;
-import org.elasticsearch.client.RequestOptions;
+import org.opensearch.action.admin.indices.settings.put.UpdateSettingsRequest;
+import org.opensearch.client.RequestOptions;
 
 import static com.linkedin.datahub.upgrade.system.elasticsearch.util.IndexUtils.INDEX_BLOCKS_WRITE_SETTING;
 import static com.linkedin.datahub.upgrade.system.elasticsearch.util.IndexUtils.getAllReindexConfigs;

diff --git a/...ain/java/com/linkedin/datahub/upgrade/system/elasticsearch/steps/BuildIndicesPreStep.java b/...ain/java/com/linkedin/datahub/upgrade/system/elasticsearch/steps/BuildIndicesPreStep.java
@@ -19,10 +19,10 @@
 import com.linkedin.metadata.shared.ElasticSearchIndexed;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
-import org.elasticsearch.ElasticsearchStatusException;
-import org.elasticsearch.action.admin.indices.settings.put.UpdateSettingsRequest;
-import org.elasticsearch.client.RequestOptions;
-import org.elasticsearch.client.indices.ResizeRequest;
+import org.opensearch.OpenSearchStatusException;
+import org.opensearch.action.admin.indices.settings.put.UpdateSettingsRequest;
+import org.opensearch.client.RequestOptions;
+import org.opensearch.client.indices.ResizeRequest;
 
 import static com.linkedin.datahub.upgrade.system.elasticsearch.util.IndexUtils.INDEX_BLOCKS_WRITE_SETTING;
 import static com.linkedin.datahub.upgrade.system.elasticsearch.util.IndexUtils.getAllReindexConfigs;
@@ -97,7 +97,7 @@ private boolean blockWrites(String indexName) throws InterruptedException, IOExc
       ack = _esComponents.getSearchClient().indices()
               .putSettings(request, RequestOptions.DEFAULT).isAcknowledged();
       log.info("Updated index {} with new settings. Settings: {}, Acknowledged: {}", indexName, indexSettings, ack);
-    } catch (ElasticsearchStatusException | IOException ese) {
+    } catch (OpenSearchStatusException | IOException ese) {
       // Cover first run case, indices won't exist so settings updates won't work nor will the rest of the preConfigure steps.
       // Since no data are in there they are skippable.
       // Have to hack around HighLevelClient not sending the actual Java type nor having an easy way to extract it :(

diff --git a/...c/main/java/com/linkedin/datahub/upgrade/system/elasticsearch/steps/CleanIndicesStep.java b/...c/main/java/com/linkedin/datahub/upgrade/system/elasticsearch/steps/CleanIndicesStep.java
@@ -9,7 +9,7 @@
 import com.linkedin.metadata.search.elasticsearch.indexbuilder.ESIndexBuilder;
 import com.linkedin.metadata.shared.ElasticSearchIndexed;
 import lombok.extern.slf4j.Slf4j;
-import org.elasticsearch.client.RestHighLevelClient;
+import org.opensearch.client.RestHighLevelClient;
 
 import java.util.List;
 import java.util.function.Function;

diff --git a/...rade/src/main/java/com/linkedin/datahub/upgrade/system/elasticsearch/util/IndexUtils.java b/...rade/src/main/java/com/linkedin/datahub/upgrade/system/elasticsearch/util/IndexUtils.java
@@ -4,12 +4,12 @@
 import com.linkedin.metadata.shared.ElasticSearchIndexed;
 import lombok.extern.slf4j.Slf4j;
 import org.apache.commons.lang3.NotImplementedException;
-import org.elasticsearch.action.admin.indices.alias.get.GetAliasesRequest;
-import org.elasticsearch.action.admin.indices.settings.get.GetSettingsRequest;
-import org.elasticsearch.action.admin.indices.settings.get.GetSettingsResponse;
-import org.elasticsearch.client.GetAliasesResponse;
-import org.elasticsearch.client.RequestOptions;
-import org.elasticsearch.client.RestHighLevelClient;
+import org.opensearch.action.admin.indices.alias.get.GetAliasesRequest;
+import org.opensearch.action.admin.indices.settings.get.GetSettingsRequest;
+import org.opensearch.action.admin.indices.settings.get.GetSettingsResponse;
+import org.opensearch.client.GetAliasesResponse;
+import org.opensearch.client.RequestOptions;
+import org.opensearch.client.RestHighLevelClient;
 
 import java.io.IOException;
 import java.util.ArrayList;

diff --git a/docker/build.gradle b/docker/build.gradle
@@ -38,6 +38,16 @@ task quickstart(type: Exec, dependsOn: ':metadata-ingestion:install') {
     // environment "ACTIONS_VERSION", 'alpine3.17-slim'
     // environment "DATAHUB_ACTIONS_IMAGE", 'nginx'
 
+    // Elastic
+    // environment "DATAHUB_SEARCH_IMAGE", 'elasticsearch'
+    // environment "DATAHUB_SEARCH_TAG", '7.10.1'
+
+    // OpenSearch
+    environment "DATAHUB_SEARCH_IMAGE", 'opensearchproject/opensearch'
+    environment "DATAHUB_SEARCH_TAG", '2.9.0'
+    environment "XPACK_SECURITY_ENABLED", 'plugins.security.disabled=true'
+    environment "USE_AWS_ELASTICSEARCH", 'true'
+
     def cmd = [
             'source ../metadata-ingestion/venv/bin/activate && ',
             'datahub docker quickstart',

diff --git a/docker/docker-compose-with-cassandra.yml b/docker/docker-compose-with-cassandra.yml
@@ -96,6 +96,9 @@ services:
       context: ../
       dockerfile: docker/elasticsearch-setup/Dockerfile
     env_file: elasticsearch-setup/env/docker.env
+    environment:
+      - ELASTICSEARCH_USE_SSL=${ELASTICSEARCH_USE_SSL:-false}
+      - USE_AWS_ELASTICSEARCH=${USE_AWS_ELASTICSEARCH:-false}
     depends_on:
       elasticsearch:
         condition: service_healthy
@@ -117,13 +120,13 @@ services:
   elasticsearch:
     container_name: elasticsearch
     hostname: elasticsearch
-    image: elasticsearch:7.10.1
+    image: ${DATAHUB_SEARCH_IMAGE:-elasticsearch}:${DATAHUB_SEARCH_TAG:-7.10.1}
     ports:
       - 9200:9200
     env_file: elasticsearch/env/docker.env
     environment:
       - discovery.type=single-node
-      - xpack.security.enabled=false
+      - ${XPACK_SECURITY_ENABLED:-xpack.security.enabled=false}
     healthcheck:
       test: curl -sS --fail http://elasticsearch:9200/_cluster/health?wait_for_status=yellow&timeout=0s
       start_period: 5s

diff --git a/docker/docker-compose-without-neo4j.yml b/docker/docker-compose-without-neo4j.yml
@@ -81,6 +81,9 @@ services:
       context: ../
       dockerfile: docker/elasticsearch-setup/Dockerfile
     env_file: elasticsearch-setup/env/docker.env
+    environment:
+      - ELASTICSEARCH_USE_SSL=${ELASTICSEARCH_USE_SSL:-false}
+      - USE_AWS_ELASTICSEARCH=${USE_AWS_ELASTICSEARCH:-false}
     depends_on:
       elasticsearch:
         condition: service_healthy
@@ -104,13 +107,13 @@ services:
   elasticsearch:
     container_name: elasticsearch
     hostname: elasticsearch
-    image: elasticsearch:7.10.1
+    image: ${DATAHUB_SEARCH_IMAGE:-elasticsearch}:${DATAHUB_SEARCH_TAG:-7.10.1}
     ports:
     - ${DATAHUB_MAPPED_ELASTIC_PORT:-9200}:9200
     env_file: elasticsearch/env/docker.env
     environment:
     - discovery.type=single-node
-    - xpack.security.enabled=false
+    - ${XPACK_SECURITY_ENABLED:-xpack.security.enabled=false}
     deploy:
       resources:
         limits: