---
subcategory: "Bedrock Agents"
layout: "aws"
page_title: "AWS: aws_bedrockagent_knowledge_base"
description: |-
  Manages an AWS Agents for Amazon Bedrock Knowledge Base.
---

# Resource: aws_bedrockagent_knowledge_base

Manages an AWS Agents for Amazon Bedrock Knowledge Base.

## Example Usage

### Basic Usage

```terraform
resource "aws_bedrockagent_knowledge_base" "example" {
  name     = "example"
  role_arn = aws_iam_role.example.arn
  knowledge_base_configuration {
    vector_knowledge_base_configuration {
      embedding_model_arn = "arn:aws:bedrock:us-west-2::foundation-model/amazon.titan-embed-text-v2:0"
    }
    type = "VECTOR"
  }
  storage_configuration {
    type = "OPENSEARCH_SERVERLESS"
    opensearch_serverless_configuration {
      collection_arn    = "arn:aws:aoss:us-west-2:123456789012:collection/142bezjddq707i5stcrf"
      vector_index_name = "bedrock-knowledge-base-default-index"
      field_mapping {
        vector_field   = "bedrock-knowledge-base-default-vector"
        text_field     = "AMAZON_BEDROCK_TEXT_CHUNK"
        metadata_field = "AMAZON_BEDROCK_METADATA"
      }
    }
  }
}
```

### Kendra Knowledge Base

```terraform
resource "aws_bedrockagent_knowledge_base" "kendra_example" {
  name     = "example-kendra-kb"
  role_arn = aws_iam_role.example.arn

  knowledge_base_configuration {
    type = "KENDRA"
    kendra_knowledge_base_configuration {
      kendra_index_arn = "arn:aws:kendra:us-east-1:123456789012:index/example-index-id"
    }
  }
}
```

### Structured Data Store

```terraform
resource "aws_bedrockagent_knowledge_base" "example" {
  name     = "example-kb"
  role_arn = aws_iam_role.example.arn

  knowledge_base_configuration {
    type = "SQL"

    sql_knowledge_base_configuration {
      type = "REDSHIFT"

      redshift_configuration {
        query_engine_configuration {
          type = "PROVISIONED"

          provisioned_configuration {
            cluster_identifier = aws_redshift_cluster.example.cluster_identifier

            auth_configuration {
              type          = "USERNAME"
              database_user = aws_redshift_cluster.example.master_username
            }
          }
        }

        storage_configuration {
          type = "REDSHIFT"

          redshift_configuration {
            database_name = aws_redshift_cluster.example.database_name
          }
        }
      }
    }
  }
}
```

### OpenSearch Managed Cluster Configuration

```terraform
resource "aws_bedrockagent_knowledge_base" "example" {
  name     = "example"
  role_arn = aws_iam_role.example.arn

  knowledge_base_configuration {
    vector_knowledge_base_configuration {
      embedding_model_arn = "arn:aws:bedrock:us-west-2::foundation-model/amazon.titan-embed-text-v2:0"
    }
    type = "VECTOR"
  }

  storage_configuration {
    type = "OPENSEARCH_MANAGED_CLUSTER"
    opensearch_managed_cluster_configuration {
      domain_arn        = "arn:aws:es:us-west-2:123456789012:domain/example-domain"
      domain_endpoint   = "https://search-example-domain.us-west-2.es.amazonaws.com"
      vector_index_name = "example_index"

      field_mapping {
        metadata_field = "metadata"
        text_field     = "chunks"
        vector_field   = "embedding"
      }
    }
  }
}
```

### With Supplemental Data Storage Configuration

```terraform
resource "aws_bedrockagent_knowledge_base" "example" {
  name     = "example"
  role_arn = aws_iam_role.example.arn
  knowledge_base_configuration {
    vector_knowledge_base_configuration {
      embedding_model_arn = "arn:aws:bedrock:us-west-2::foundation-model/amazon.titan-embed-text-v2:0"

      embedding_model_configuration {
        bedrock_embedding_model_configuration {
          dimensions          = 1024
          embedding_data_type = "FLOAT32"

          audio {
            segmentation_configuration {
              fixed_length_duration = 60
            }
          }

          video {
            segmentation_configuration {
              fixed_length_duration = 60
            }
          }
        }
      }

      supplemental_data_storage_configuration {
        storage_location {
          type = "S3"

          s3_location {
            uri = "s3://my-bucket/chunk-processor/"
          }
        }
      }
    }
    type = "VECTOR"
  }

  storage_configuration {
    type = "OPENSEARCH_SERVERLESS"
    opensearch_serverless_configuration {
      collection_arn    = "arn:aws:aoss:us-west-2:123456789012:collection/142bezjddq707i5stcrf"
      vector_index_name = "bedrock-knowledge-base-default-index"
      field_mapping {
        vector_field   = "bedrock-knowledge-base-default-vector"
        text_field     = "AMAZON_BEDROCK_TEXT_CHUNK"
        metadata_field = "AMAZON_BEDROCK_METADATA"
      }
    }
  }
}
```

### S3 Vectors Configuration

```terraform
resource "aws_s3vectors_vector_bucket" "example" {
  vector_bucket_name = "example-bucket"
}

resource "aws_s3vectors_index" "example" {
  index_name         = "example-index"
  vector_bucket_name = aws_s3vectors_vector_bucket.example.vector_bucket_name

  data_type       = "float32"
  dimension       = 256
  distance_metric = "euclidean"
}

resource "aws_bedrockagent_knowledge_base" "example" {
  name     = "example-s3vectors-kb"
  role_arn = aws_iam_role.example.arn

  knowledge_base_configuration {
    vector_knowledge_base_configuration {
      embedding_model_arn = "arn:aws:bedrock:us-west-2::foundation-model/amazon.titan-embed-text-v2:0"
      embedding_model_configuration {
        bedrock_embedding_model_configuration {
          dimensions          = 256
          embedding_data_type = "FLOAT32"
        }
      }
    }
    type = "VECTOR"
  }

  storage_configuration {
    type = "S3_VECTORS"
    s3_vectors_configuration {
      index_arn = aws_s3vectors_index.example.index_arn
    }
  }
}
```

### Managed Knowledge Base

```terraform
resource "aws_bedrockagent_knowledge_base" "example" {
  name     = "example-managed-kb"
  role_arn = aws_iam_role.example.arn

  knowledge_base_configuration {
    type = "MANAGED"

    managed_knowledge_base_configuration {
      embedding_model_type = "MANAGED"
    }
  }
}
```

### Managed Knowledge Base with Custom Embedding Model

```terraform
resource "aws_bedrockagent_knowledge_base" "example" {
  name     = "example-managed-multilingual-kb"
  role_arn = aws_iam_role.example.arn

  knowledge_base_configuration {
    type = "MANAGED"

    managed_knowledge_base_configuration {
      embedding_model_type = "CUSTOM"
      embedding_model_arn  = "arn:aws:bedrock:us-east-1::foundation-model/cohere.embed-multilingual-v3"

      embedding_model_configuration {
        bedrock_embedding_model_configuration {
          dimensions = 1024
        }
      }
    }
  }
}
```

## Argument Reference

The following arguments are required:

* `knowledge_base_configuration` - (Required, Forces new resource) Details about the embeddings configuration of the knowledge base. See [`knowledge_base_configuration` Block](#knowledge_base_configuration-block) for details.
* `name` - (Required) Name of the knowledge base.
* `role_arn` - (Required) ARN of the IAM role with permissions to invoke API operations on the knowledge base.

The following arguments are optional:

* `description` - (Optional) Description of the knowledge base.
* `region` - (Optional) Region where this resource will be [managed](https://docs.aws.amazon.com/general/latest/gr/rande.html#regional-endpoints). Defaults to the Region set in the [provider configuration](https://registry.terraform.io/providers/hashicorp/aws/latest/docs#aws-configuration-reference).
* `storage_configuration` - (Optional, Forces new resource) Details about the storage configuration of the knowledge base. See [`storage_configuration` Block](#storage_configuration-block) for details.
* `tags` - (Optional) Map of tags assigned to the resource. If configured with a provider [`default_tags` configuration block](/docs/providers/aws/index.html#default_tags-configuration-block) present, tags with matching keys will overwrite those defined at the provider-level.

### `knowledge_base_configuration` Block

The `knowledge_base_configuration` configuration block supports the following arguments:

* `kendra_knowledge_base_configuration` - (Optional) Settings for an Amazon Kendra knowledge base. See [`kendra_knowledge_base_configuration` Block](#kendra_knowledge_base_configuration-block) for details.
* `managed_knowledge_base_configuration` - (Optional) Settings for a managed knowledge base where Amazon Bedrock manages the vector store. See [`managed_knowledge_base_configuration` Block](#managed_knowledge_base_configuration-block) for details.
* `sql_knowledge_base_configuration` - (Optional) Configurations for a knowledge base connected to an SQL database. See [`sql_knowledge_base_configuration` Block](#sql_knowledge_base_configuration-block) for details.
* `type` - (Required) Type of data that the data source is converted into for the knowledge base. Valid Values: `VECTOR`, `KENDRA`, `SQL`, `MANAGED`.
* `vector_knowledge_base_configuration` - (Optional) Details about the model that's used to convert the data source into vector embeddings. See [`vector_knowledge_base_configuration` Block](#vector_knowledge_base_configuration-block) for details.

### `kendra_knowledge_base_configuration` Block

The `kendra_knowledge_base_configuration` configuration block supports the following arguments:

* `kendra_index_arn` - (Required) ARN of the Amazon Kendra index.

### `managed_knowledge_base_configuration` Block

The `managed_knowledge_base_configuration` configuration block supports the following arguments:

* `embedding_model_arn` - (Optional) ARN of the embedding model. Required when `embedding_model_type` is `CUSTOM`.
* `embedding_model_configuration` - (Optional) Configuration for the embedding model. Required when `embedding_model_type` is `CUSTOM`. See [`embedding_model_configuration` Block](#embedding_model_configuration-block) for details.
* `embedding_model_type` - (Optional) Type of embedding model. Valid values: `MANAGED`, `CUSTOM`. When `MANAGED`, no model selection or configuration is required. When `CUSTOM`, `embedding_model_arn` and `embedding_model_configuration` are required. Defaults to `MANAGED`.
* `server_side_encryption_configuration` - (Optional) Server-side encryption configuration for the managed knowledge base. See [`server_side_encryption_configuration` Block](#server_side_encryption_configuration-block) for details.

### `server_side_encryption_configuration` Block

The `server_side_encryption_configuration` configuration block supports the following arguments:

* `kms_key_arn` - (Optional) ARN of the KMS key used to encrypt the managed knowledge base.

~> **Note:** `storage_configuration` is not required when `knowledge_base_configuration.type` is `MANAGED`. Amazon Bedrock manages the vector store automatically for managed knowledge bases.

### `sql_knowledge_base_configuration` Block

The `sql_knowledge_base_configuration` configuration block supports the following arguments:

* `redshift_configuration` - (Required) Configurations for a knowledge base connected to an Amazon Redshift database. See [`knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration` Block](#knowledge_base_configurationsql_knowledge_base_configurationredshift_configuration-block) for details.
* `type` - (Required) Type of SQL database to connect to the knowledge base. Valid values: `REDSHIFT`.

### `knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration` Block

The `redshift_configuration` configuration block supports the following arguments:

* `query_engine_configuration` - (Required) Configurations for an Amazon Redshift query engine. See [`query_engine_configuration` Block](#query_engine_configuration-block) for details.
* `query_generation_configuration` - (Optional) Configurations for generating queries. See [`query_generation_configuration` Block](#query_generation_configuration-block) for details.
* `storage_configuration` - (Required) Configurations for Amazon Redshift database storage. See [`knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.storage_configuration` Block](#knowledge_base_configurationsql_knowledge_base_configurationredshift_configurationstorage_configuration-block) for details.

### `query_engine_configuration` Block

The `query_engine_configuration` configuration block supports the following arguments:

* `provisioned_configuration` - (Optional) Configurations for a provisioned Amazon Redshift query engine. See [`provisioned_configuration` Block](#provisioned_configuration-block) for details.
* `serverless_configuration` - (Optional) Configurations for a serverless Amazon Redshift query engine. See [`serverless_configuration` Block](#serverless_configuration-block) for details.
* `type` - (Required) Type of query engine. Valid values: `SERVERLESS`, `PROVISIONED`.

### `provisioned_configuration` Block

The `provisioned_configuration` configuration block supports the following arguments:

* `auth_configuration` - (Required) Configurations for authentication to Amazon Redshift. See [`knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.query_engine_configuration.provisioned_configuration.auth_configuration` Block](#knowledge_base_configurationsql_knowledge_base_configurationredshift_configurationquery_engine_configurationprovisioned_configurationauth_configuration-block) for details.
* `cluster_identifier` - (Required) ID of the Amazon Redshift cluster.

### `knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.query_engine_configuration.provisioned_configuration.auth_configuration` Block

The `auth_configuration` configuration block supports the following arguments:

* `database_user` - (Optional) Database username for authentication to an Amazon Redshift provisioned data warehouse.
* `type` - (Required) Type of authentication to use. Valid values: `IAM`, `USERNAME_PASSWORD`, `USERNAME`.
* `username_password_secret_arn` - (Optional) ARN of a Secrets Manager secret for authentication.

### `serverless_configuration` Block

The `serverless_configuration` configuration block supports the following arguments:

* `auth_configuration` - (Required) Configurations for authentication to Amazon Redshift Serverless. See [`knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.query_engine_configuration.serverless_configuration.auth_configuration` Block](#knowledge_base_configurationsql_knowledge_base_configurationredshift_configurationquery_engine_configurationserverless_configurationauth_configuration-block) for details.
* `workgroup_arn` - (Required) ARN of the Amazon Redshift workgroup.

### `knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.query_engine_configuration.serverless_configuration.auth_configuration` Block

The `auth_configuration` configuration block supports the following arguments:

* `type` - (Required) Type of authentication to use. Valid values: `IAM`, `USERNAME_PASSWORD`.
* `username_password_secret_arn` - (Optional) ARN of a Secrets Manager secret for authentication.

### `query_generation_configuration` Block

The `query_generation_configuration` configuration block supports the following arguments:

* `execution_timeout_seconds` - (Optional) Time after which query generation will time out.
* `generation_context` - (Optional) Configurations for context to use during query generation. See [`generation_context` Block](#generation_context-block) for details.

### `generation_context` Block

The `generation_context` configuration block supports the following arguments:

* `curated_query` - (Optional) Information about example queries to help the query engine generate appropriate SQL queries. See [`curated_query` Block](#curated_query-block) for details.
* `table` - (Optional) Information about a table in the database. See [`table` Block](#table-block) for details.

### `curated_query` Block

The `curated_query` configuration block supports the following arguments:

* `natural_language` - (Required) Example natural language query.
* `sql` - (Required) SQL equivalent of `natural_language`.

### `table` Block

The `table` configuration block supports the following arguments:

* `column` - (Optional) Information about a column in the table. See [`column` Block](#column-block) for details.
* `description` - (Optional) Description of the table that helps the query engine understand the contents of the table.
* `inclusion` - (Optional) Whether to include or exclude the table during query generation. Valid values `INCLUDE`, `EXCLUDE`.
* `name` - (Required) Name of the table for which the other fields in this object apply.

### `column` Block

The `column` configuration block supports the following arguments:

* `description` - (Optional) Description of the column that helps the query engine understand the contents of the column.
* `inclusion` - (Optional) Whether to include or exclude the column during query generation. Valid values `INCLUDE`, `EXCLUDE`.
* `name` - (Optional) Name of the column for which the other fields in this object apply.

### `knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.storage_configuration` Block

The `storage_configuration` configuration block supports the following arguments:

* `aws_data_catalog_configuration` - (Optional) Configurations for storage in AWS Glue Data Catalog. See [`aws_data_catalog_configuration` Block](#aws_data_catalog_configuration-block) for details.
* `redshift_configuration` - (Optional) Configurations for storage in Amazon Redshift. See [`knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.storage_configuration.redshift_configuration` Block](#knowledge_base_configurationsql_knowledge_base_configurationredshift_configurationstorage_configurationredshift_configuration-block) for details.
* `type` - (Required) Data storage service to use. Valid values: `REDSHIFT`, `AWS_DATA_CATALOG`.

### `aws_data_catalog_configuration` Block

The `aws_data_catalog_configuration` configuration block supports the following arguments:

* `table_names` - (Required) List of names of the tables to use.

### `knowledge_base_configuration.sql_knowledge_base_configuration.redshift_configuration.storage_configuration.redshift_configuration` Block

The `redshift_configuration` configuration block supports the following arguments:

* `database_name` - (Required) Name of the Amazon Redshift database.

### `vector_knowledge_base_configuration` Block

The `vector_knowledge_base_configuration` configuration block supports the following arguments:

* `embedding_model_arn` - (Required) ARN of the model used to create vector embeddings for the knowledge base.
* `embedding_model_configuration` - (Optional) Embeddings model configuration details for the vector model used in the knowledge base. See [`embedding_model_configuration` Block](#embedding_model_configuration-block) for details.
* `supplemental_data_storage_configuration` - (Optional) Supplemental data storage configuration for images extracted from multimodal documents. See [`supplemental_data_storage_configuration` Block](#supplemental_data_storage_configuration-block) for details.

### `embedding_model_configuration` Block

The `embedding_model_configuration` configuration block supports the following arguments:

* `bedrock_embedding_model_configuration` - (Optional) Vector configuration details for the Bedrock embeddings model. See [`bedrock_embedding_model_configuration` Block](#bedrock_embedding_model_configuration-block) for details.

### `bedrock_embedding_model_configuration` Block

The `bedrock_embedding_model_configuration` configuration block supports the following arguments:

* `audio` - (Optional) Configuration for processing audio content in multimodal knowledge bases. See [`audio` Block](#audio-block) for details.
* `dimensions` - (Optional) Dimension details for the vector configuration used on the Bedrock embeddings model.
* `embedding_data_type` - (Optional) Data type for the vectors when using a model to convert text into vector embeddings. The model must support the specified data type for vector embeddings. Valid values are `FLOAT32` and `BINARY`.
* `video` - (Optional) Configuration for processing video content in multimodal knowledge bases. See [`video` Block](#video-block) for details.

### `audio` Block

The `audio` configuration block supports the following arguments:

* `segmentation_configuration` - (Required) Configuration for segmenting audio content during processing. See [`segmentation_configuration` Block](#segmentation_configuration-block) for details.

### `video` Block

The `video` configuration block supports the following arguments:

* `segmentation_configuration` - (Required) Configuration for segmenting video content during processing. See [`segmentation_configuration` Block](#segmentation_configuration-block) for details.

### `segmentation_configuration` Block

The `segmentation_configuration` configuration block supports the following arguments:

* `fixed_length_duration` - (Required) Duration in seconds for each audio or video segment.

### `supplemental_data_storage_configuration` Block

The `supplemental_data_storage_configuration` configuration block supports the following arguments:

* `storage_location` - (Required) Storage location specification for images extracted from multimodal documents in your data source. See [`storage_location` Block](#storage_location-block) for details.

### `storage_location` Block

The `storage_location` configuration block supports the following arguments:

* `s3_location` - (Required) Information about the Amazon S3 location for the extracted images. See [`s3_location` Block](#s3_location-block) for details.
* `type` - (Required) Storage service used for this location. `S3` is the only valid value.

### `s3_location` Block

The `s3_location` configuration block supports the following arguments:

* `uri` - (Required) URI of the location.

### `storage_configuration` Block

The `storage_configuration` configuration block supports the following arguments:

* `mongo_db_atlas_configuration` - (Optional) Storage configuration of the knowledge base in MongoDB Atlas. See [`mongo_db_atlas_configuration` Block](#mongo_db_atlas_configuration-block) for details.
* `neptune_analytics_configuration` - (Optional) Storage configuration of the knowledge base in Amazon Neptune Analytics. See [`neptune_analytics_configuration` Block](#neptune_analytics_configuration-block) for details.
* `opensearch_managed_cluster_configuration` - (Optional) Storage configuration of the knowledge base in Amazon OpenSearch Service Managed Cluster. See [`opensearch_managed_cluster_configuration` Block](#opensearch_managed_cluster_configuration-block) for details.
* `opensearch_serverless_configuration` - (Optional) Storage configuration of the knowledge base in Amazon OpenSearch Service Serverless. See [`opensearch_serverless_configuration` Block](#opensearch_serverless_configuration-block) for details.
* `pinecone_configuration` - (Optional) Storage configuration of the knowledge base in Pinecone. See [`pinecone_configuration` Block](#pinecone_configuration-block) for details.
* `rds_configuration` - (Optional) Details about the storage configuration of the knowledge base in Amazon RDS. For more information, see [Create a vector index in Amazon RDS](https://docs.aws.amazon.com/bedrock/latest/userguide/knowledge-base-setup.html). See [`rds_configuration` Block](#rds_configuration-block) for details.
* `redis_enterprise_cloud_configuration` - (Optional) Storage configuration of the knowledge base in Redis Enterprise Cloud. See [`redis_enterprise_cloud_configuration` Block](#redis_enterprise_cloud_configuration-block) for details.
* `s3_vectors_configuration` - (Optional) Storage configuration of the knowledge base in Amazon S3 Vectors. See [`s3_vectors_configuration` Block](#s3_vectors_configuration-block) for details.
* `type` - (Required) Vector store service in which the knowledge base is stored. Valid Values: `MONGO_DB_ATLAS`, `OPENSEARCH_SERVERLESS`, `OPENSEARCH_MANAGED_CLUSTER`, `PINECONE`, `REDIS_ENTERPRISE_CLOUD`, `RDS`, `S3_VECTORS`, `NEPTUNE_ANALYTICS`.

### `mongo_db_atlas_configuration` Block

The `mongo_db_atlas_configuration` configuration block supports the following arguments:

* `collection_name` - (Required) Name of the collection in the MongoDB Atlas database.
* `credentials_secret_arn` - (Required) ARN of the secret that you created in AWS Secrets Manager that is linked to your MongoDB Atlas database.
* `database_name` - (Required) Name of the database in the MongoDB Atlas database.
* `endpoint` - (Required) Endpoint URL of the MongoDB Atlas database.
* `endpoint_service_name` - (Optional) Name of the VPC endpoint service in your account that is connected to your MongoDB Atlas cluster.
* `field_mapping` - (Required) Names of the fields to which to map information about the vector store. See [`storage_configuration.mongo_db_atlas_configuration.field_mapping` Block](#storage_configurationmongo_db_atlas_configurationfield_mapping-block) for details.
* `text_index_name` - (Optional) Name of the text search index in the MongoDB collection. Required for hybrid search.
* `vector_index_name` - (Required) Name of the vector index.

### `storage_configuration.mongo_db_atlas_configuration.field_mapping` Block

The `field_mapping` configuration block supports the following arguments:

* `metadata_field` - (Required) Name of the field in which Amazon Bedrock stores metadata about the vector store.
* `text_field` - (Required) Name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.
* `vector_field` - (Required) Name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

### `neptune_analytics_configuration` Block

The `neptune_analytics_configuration` configuration block supports the following arguments:

* `field_mapping` - (Required) Names of the fields to which to map information about the vector store. See [`storage_configuration.neptune_analytics_configuration.field_mapping` Block](#storage_configurationneptune_analytics_configurationfield_mapping-block) for details.
* `graph_arn` - (Required) ARN of the Neptune Analytics vector store.

### `storage_configuration.neptune_analytics_configuration.field_mapping` Block

The `field_mapping` configuration block supports the following arguments:

* `metadata_field` - (Required) Name of the field in which Amazon Bedrock stores metadata about the vector store.
* `text_field` - (Required) Name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

### `opensearch_managed_cluster_configuration` Block

The `opensearch_managed_cluster_configuration` configuration block supports the following arguments:

* `domain_arn` - (Required) ARN of the OpenSearch domain.
* `domain_endpoint` - (Required) Endpoint URL of the OpenSearch domain.
* `field_mapping` - (Required) Names of the fields to which to map information about the vector store. See [`storage_configuration.opensearch_managed_cluster_configuration.field_mapping` Block](#storage_configurationopensearch_managed_cluster_configurationfield_mapping-block) for details.
* `vector_index_name` - (Required) Name of the vector store.

### `storage_configuration.opensearch_managed_cluster_configuration.field_mapping` Block

The `field_mapping` configuration block supports the following arguments:

* `metadata_field` - (Required) Name of the field in which Amazon Bedrock stores metadata about the vector store.
* `text_field` - (Required) Name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.
* `vector_field` - (Required) Name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

### `opensearch_serverless_configuration` Block

The `opensearch_serverless_configuration` configuration block supports the following arguments:

* `collection_arn` - (Required) ARN of the OpenSearch Service vector store.
* `field_mapping` - (Required) Names of the fields to which to map information about the vector store. See [`storage_configuration.opensearch_serverless_configuration.field_mapping` Block](#storage_configurationopensearch_serverless_configurationfield_mapping-block) for details.
* `vector_index_name` - (Required) Name of the vector store.

### `storage_configuration.opensearch_serverless_configuration.field_mapping` Block

The `field_mapping` configuration block supports the following arguments:

* `metadata_field` - (Required) Name of the field in which Amazon Bedrock stores metadata about the vector store.
* `text_field` - (Required) Name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.
* `vector_field` - (Required) Name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

### `pinecone_configuration` Block

The `pinecone_configuration` configuration block supports the following arguments:

* `connection_string` - (Required) Endpoint URL for your index management page.
* `credentials_secret_arn` - (Required) ARN of the secret that you created in AWS Secrets Manager that is linked to your Pinecone API key.
* `field_mapping` - (Required) Names of the fields to which to map information about the vector store. See [`storage_configuration.pinecone_configuration.field_mapping` Block](#storage_configurationpinecone_configurationfield_mapping-block) for details.
* `namespace` - (Optional) Namespace to be used to write new data to your database.

### `storage_configuration.pinecone_configuration.field_mapping` Block

The `field_mapping` configuration block supports the following arguments:

* `metadata_field` - (Required) Name of the field in which Amazon Bedrock stores metadata about the vector store.
* `text_field` - (Required) Name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.

### `rds_configuration` Block

The `rds_configuration` configuration block supports the following arguments:

* `credentials_secret_arn` - (Required) ARN of the secret that you created in AWS Secrets Manager that is linked to your Amazon RDS database.
* `database_name` - (Required) Name of your Amazon RDS database.
* `field_mapping` - (Required) Names of the fields to which to map information about the vector store. See [`storage_configuration.rds_configuration.field_mapping` Block](#storage_configurationrds_configurationfield_mapping-block) for details.
* `resource_arn` - (Required) ARN of the vector store.
* `table_name` - (Required) Name of the table in the database.

### `storage_configuration.rds_configuration.field_mapping` Block

The `field_mapping` configuration block supports the following arguments:

* `custom_metadata_field` - (Optional) Name for the universal metadata field where Amazon Bedrock will store any custom metadata from your data source.
* `metadata_field` - (Required) Name of the field in which Amazon Bedrock stores metadata about the vector store.
* `primary_key_field` - (Required) Name of the field in which Amazon Bedrock stores the ID for each entry.
* `text_field` - (Required) Name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.
* `vector_field` - (Required) Name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

### `redis_enterprise_cloud_configuration` Block

The `redis_enterprise_cloud_configuration` configuration block supports the following arguments:

* `credentials_secret_arn` - (Required) ARN of the secret that you created in AWS Secrets Manager that is linked to your Redis Enterprise Cloud database.
* `endpoint` - (Required) Endpoint URL of the Redis Enterprise Cloud database.
* `field_mapping` - (Required) Names of the fields to which to map information about the vector store. See [`storage_configuration.redis_enterprise_cloud_configuration.field_mapping` Block](#storage_configurationredis_enterprise_cloud_configurationfield_mapping-block) for details.
* `vector_index_name` - (Required) Name of the vector index.

### `storage_configuration.redis_enterprise_cloud_configuration.field_mapping` Block

The `field_mapping` configuration block supports the following arguments:

* `metadata_field` - (Optional) Name of the field in which Amazon Bedrock stores metadata about the vector store.
* `text_field` - (Optional) Name of the field in which Amazon Bedrock stores the raw text from your data. The text is split according to the chunking strategy you choose.
* `vector_field` - (Optional) Name of the field in which Amazon Bedrock stores the vector embeddings for your data sources.

### `s3_vectors_configuration` Block

The `s3_vectors_configuration` configuration block supports the following arguments.
Either `index_arn`, or both `index_name` and `vector_bucket_arn` must be specified.

* `index_arn` - (Optional) ARN of the S3 Vectors index. Conflicts with `index_name` and `vector_bucket_arn`.
* `index_name` - (Optional) Name of the S3 Vectors index. Must be specified with `vector_bucket_arn`. Conflicts with `index_arn`.
* `vector_bucket_arn` - (Optional) ARN of the S3 Vectors vector bucket. Must be specified with `index_name`. Conflicts with `index_arn`.

## Attribute Reference

This resource exports the following attributes in addition to the arguments above:

* `arn` - ARN of the knowledge base.
* `created_at` - Time at which the knowledge base was created.
* `failure_reasons` - List of failure reasons reported when the knowledge base is in a failed state.
* `id` - Unique identifier of the knowledge base.
* `tags_all` - Map of tags assigned to the resource, including those inherited from the provider [`default_tags` configuration block](https://registry.terraform.io/providers/hashicorp/aws/latest/docs#default_tags-configuration-block).
* `updated_at` - Time at which the knowledge base was last updated.

## Timeouts

[Configuration options](https://developer.hashicorp.com/terraform/language/resources/syntax#operation-timeouts):

* `create` - (Default `30m`)
* `update` - (Default `30m`)
* `delete` - (Default `30m`)

## Import

In Terraform v1.5.0 and later, use an [`import` block](https://developer.hashicorp.com/terraform/language/import) to import Agents for Amazon Bedrock Knowledge Base using the knowledge base ID. For example:

```terraform
import {
  to = aws_bedrockagent_knowledge_base.example
  id = "EMDPPAYPZI"
}
```

Using `terraform import`, import Agents for Amazon Bedrock Knowledge Base using the knowledge base ID. For example:

```console
% terraform import aws_bedrockagent_knowledge_base.example EMDPPAYPZI
```