feat(schema): remove embedding fields from naturalLanguageSearch (moved to llmConfiguration.embeddings)

Remove dead embeddingProvider, maxConcurrentRequests, bedrock, openai, google, djl
properties from naturalLanguageSearch in elasticSearchConfiguration.json — these
moved to llmConfiguration.embeddings in prior commits and OSS code no longer reads
them from the old location.  Regenerate NaturalLanguageSearchConfiguration POJO,
drop the no-longer-generated Bedrock/Djl/Openai/Google inner classes, and remove
two leftover test references (stale mock setup in SystemRepositoryEmbeddingsValidationTest,
dead withEmbeddingProvider() call in SearchRepositoryBehaviorTest).

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Author: pmbrull

openmetadata-service/src/test/java/org/openmetadata/service/jdbi3/SystemRepositoryEmbeddingsValidationTest.java

@@ -12,7 +12,6 @@
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.Test;
 import org.mockito.MockedStatic;
-import org.openmetadata.schema.service.configuration.elasticsearch.Bedrock;
 import org.openmetadata.schema.service.configuration.elasticsearch.ElasticSearchConfiguration;
 import org.openmetadata.schema.service.configuration.elasticsearch.NaturalLanguageSearchConfiguration;
 import org.openmetadata.schema.system.StepValidation;
@@ -53,9 +52,6 @@ void setup() {
     NaturalLanguageSearchConfiguration nlpConfig = mock(NaturalLanguageSearchConfiguration.class);
     when(appConfig.getElasticSearchConfiguration()).thenReturn(esConfig);
     when(esConfig.getNaturalLanguageSearch()).thenReturn(nlpConfig);
-    when(nlpConfig.getEmbeddingProvider()).thenReturn("bedrock");
-    Bedrock bedrockConfig = mock(Bedrock.class);
-    when(nlpConfig.getBedrock()).thenReturn(bedrockConfig);
 
     when(searchRepository.getSearchType())
         .thenReturn(ElasticSearchConfiguration.SearchType.OPENSEARCH);

openmetadata-service/src/test/java/org/openmetadata/service/search/SearchRepositoryBehaviorTest.java

@@ -2543,8 +2543,7 @@ void initializeVectorSearchServiceSkipsWhenEmbeddingsAreDisabled() {
 
   @Test
   void initializeVectorSearchServiceInitializesOpenSearchVectorSupport() throws Exception {
-    NaturalLanguageSearchConfiguration nlConfig =
-        new NaturalLanguageSearchConfiguration().withEmbeddingProvider("openai");
+    NaturalLanguageSearchConfiguration nlConfig = new NaturalLanguageSearchConfiguration();
     SearchRepository openSearchRepository =
         newRepository(
             Map.of(Entity.TABLE, TABLE_MAPPING),

openmetadata-spec/src/main/resources/json/schema/configuration/elasticSearchConfiguration.json

@@ -148,89 +148,11 @@
           "type": "number",
           "default": 0.6
         },
-        "embeddingProvider": {
-          "description": "The provider to use for generating vector embeddings (e.g., bedrock, openai, google, djl).",
-          "type": "string",
-          "default": "bedrock"
-        },
-        "maxConcurrentRequests": {
-          "description": "Maximum number of concurrent embedding and NLQ provider requests. Controls the semaphore used to throttle calls to the providers and prevent overwhelming HTTP/2 connection limits.",
-          "type": "integer",
-          "default": 10,
-          "minimum": 1
-        },
         "providerClass": {
           "description": "Fully qualified class name of the NLQService implementation to use",
           "type": "string",
           "default": "org.openmetadata.service.search.nlq.NoOpNLQService"
         },
-        "bedrock": {
-          "description": "AWS Bedrock configuration for natural language processing",
-          "type": "object",
-          "properties": {
-            "awsConfig": {
-              "description": "AWS credentials configuration for Bedrock service",
-              "$ref": "../security/credentials/awsBaseConfig.json"
-            },
-            "embeddingModelId": {
-              "description": "Bedrock embedding model identifier to use for vector search",
-              "type": "string",
-              "default": "amazon.titan-embed-text-v2:0"
-            },
-            "embeddingDimension": {
-              "description": "Dimension of the embedding vector",
-              "type": "integer",
-              "default": 512
-            }
-          },
-          "additionalProperties": false
-        },
-        "djl": {
-          "description": "Embedding generation using Deep Java Library (DJL)",
-          "type": "object",
-          "properties": {
-            "embeddingModel": {
-              "description": "DJL model name for embedding generation",
-              "type": "string"
-            }
-          },
-          "additionalProperties": false
-        },
-        "openai": {
-          "description": "OpenAI configuration for embedding generation. Supports both OpenAI and Azure OpenAI endpoints.",
-          "type": "object",
-          "javaType": "org.openmetadata.schema.service.configuration.elasticsearch.Openai",
-          "properties": {
-            "apiKey": {
-              "description": "API key for authenticating with OpenAI or Azure OpenAI.",
-              "type": "string"
-            },
-            "embeddingModelId": {
-              "description": "OpenAI embedding model identifier (e.g., text-embedding-3-small, text-embedding-ada-002).",
-              "type": "string",
-              "default": "text-embedding-3-small"
-            },
-            "embeddingDimension": {
-              "description": "Dimension of the embedding vector. Default is 1536 for text-embedding-3-small.",
-              "type": "integer",
-              "default": 1536
-            },
-            "endpoint": {
-              "description": "Custom endpoint URL. For Azure OpenAI, use the Azure resource endpoint (e.g., https://your-resource.openai.azure.com). Leave empty for standard OpenAI API.",
-              "type": "string"
-            },
-            "deploymentName": {
-              "description": "Azure OpenAI deployment name. Required when using Azure OpenAI.",
-              "type": "string"
-            },
-            "apiVersion": {
-              "description": "Azure OpenAI API version. Only used with Azure OpenAI.",
-              "type": "string",
-              "default": "2024-02-01"
-            }
-          },
-          "additionalProperties": false
-        },
         "filterExtractor": {
           "description": "NLQ filter extractor cache and prompt tuning.",
           "type": "object",
@@ -317,32 +239,6 @@
             }
           },
           "additionalProperties": false
-        },
-        "google": {
-          "description": "Google Gemini configuration for embedding generation via the Generative Language API.",
-          "type": "object",
-          "javaType": "org.openmetadata.schema.service.configuration.elasticsearch.Google",
-          "properties": {
-            "apiKey": {
-              "description": "API key from Google AI Studio for authenticating with the Generative Language API.",
-              "type": "string"
-            },
-            "embeddingModelId": {
-              "description": "Gemini embedding model identifier (e.g., gemini-embedding-001, text-embedding-004).",
-              "type": "string",
-              "default": "gemini-embedding-001"
-            },
-            "embeddingDimension": {
-              "description": "Dimension of the embedding vector, sent to Google as `outputDimensionality`. For `gemini-embedding-001` valid values are 768, 1536, or 3072. For `text-embedding-004` use 768.",
-              "type": "integer",
-              "default": 768
-            },
-            "endpoint": {
-              "description": "Optional override for the full embedding endpoint URL. Must be the complete URL including the model and `:embedContent` action (e.g. `https://generativelanguage.googleapis.com/v1beta/models/text-embedding-004:embedContent`), not just a base URL. Leave empty to use the default Generative Language API endpoint, which is constructed from `embeddingModelId`. The `key` query parameter is appended automatically.",
-              "type": "string"
-            }
-          },
-          "additionalProperties": false
         }
       },
       "additionalProperties": false