Creating collections
You can use the console or the AWS CLI to create a serverless collection. These steps cover how to create a search or time series collection. To create a vector search collection, see Working with vector search collections.
Before you create a collection, ensure that you have the required permissions. For more information, see Configuring permissions for collections.
Amazon OpenSearch Serverless supports two collection generations:
-
NextGen – The latest generation of OpenSearch Serverless with instant auto scaling and scale-to-zero for cost optimization. Uses collection groups for shared capacity management across multiple collections. Offers a simplified single-page creation flow with Express Create and Standard Create options.
-
Classic – Uses a multi-step wizard with per-collection security, network, and encryption configuration. Each collection is configured independently.
When you choose Create collection from the Collections page, the console opens the NextGen creation form by default. You can switch between generations at any time using the Switch to Classic or Switch to NextGen link in the Serverless generation field on the creation form.
Configure NextGen collection settings
The following steps are common to both Express Create and Standard Create methods.
To configure NextGen collection settings
-
Navigate to the Amazon OpenSearch Service console at https://console.aws.amazon.com/aos/home/
. -
Expand Serverless in the left navigation pane and choose Collections.
-
Choose Create collection. The NextGen collection creation form appears by default.
Tip
To create a Classic collection instead, choose Switch to Classic in the Serverless generation field.
-
Provide a name and description for the collection. The name must meet the following criteria:
-
Is unique to your account and AWS Region
-
Contains only lowercase letters a-z, the numbers 0–9, and the hyphen (-)
-
Contains between 3 and 32 characters
-
-
Choose a collection type:
-
Search – Full-text search that powers applications in your internal networks and internet-facing applications. All search data is stored in hot storage to ensure fast query response times.
-
Vector search – Semantic search on vector embeddings that simplifies vector data management. Powers machine learning (ML) augmented search experiences and generative AI applications.
For more information, see Choosing a collection type.
-
-
Choose a collection creation method: Express Create or Standard Create.
Create a NextGen collection (Express Create)
Express Create lets you quickly set up a NextGen collection with sensible defaults. OpenSearch Serverless automatically creates the required collection group, encryption, network, and data access policies for you based on the collection name and collection type you provide.
Note
You can also access Express Create from:
-
The Express create button in the blue info banner on the Amazon OpenSearch Service landing page
-
The Express create button on the Get Started Quick Create card
Both open an Express Create modal directly without navigating to the full create page.
When you use Express Create from the modal, you can create a collection in seconds without navigating to the full create collection form. Enter a collection name, choose a collection type (Search or Vector search), and choose Create collection. OpenSearch Serverless automatically configures the remaining settings with sensible defaults.
You can expand the Configuration details section to review or modify settings such as collection group, OpenSearch UI settings, encryption, network access, and data access before creating the collection. Some settings can also be changed after creation.
To create a NextGen collection using Express Create
-
After configuring the collection name, description, and type (see Configure NextGen collection settings), choose Express Create as the collection creation method.
-
Review the default settings table that OpenSearch Serverless configures on your behalf:
-
Collection group – For a first-time user, a default collection group name is auto-generated based on the collection name. You can edit the name by choosing the edit icon. For a returning user, existing NextGen collection groups of the selected type are displayed. One is selected by default. You can select any other existing collection group from the dropdown.
-
OpenSearch UI settings – Uses an existing OpenSearch application by default. The application name and workspace name are editable after creation.
-
Encryption – The collection is encrypted with an AWS owned key.
-
Network access – The collection is accessible from public networks.
-
Data access – A data access policy is automatically created that grants the current IAM principal full access to the collection and its indexes.
-
-
Choose Create collection.
The collection status shows Creating. Wait for the collection status
to change to Active before you begin indexing data. This typically
takes several minutes.
Create a NextGen collection (Standard Create)
Standard Create gives you full control over collection group, encryption, network, and data access configuration.
To create a NextGen collection using Standard Create
-
Follow steps 1–5 from Configure NextGen collection settings.
-
Under Collection creation method, select Standard create.
The form expands to show the full configuration sections described in the following sections.
Configure collection group
Choose how to assign a collection group. You can select an existing compatible group or create a new one, even if compatible groups already exist.
-
Select existing – Choose a compatible collection group from the dropdown. The group must support your selected collection type. The console displays the group's current capacity limits (min/max OCU for indexing and search).
-
Create new – Create a new collection group with custom capacity limits:
-
Collection group name – Auto-generated as
nextgen-{collection-name}. You can edit this name. The name is normalized to lowercase, hyphens replace underscores, and it is truncated to 32 characters. -
Min Indexing Capacity (in OCUs) – Optional. Leave blank for no minimum.
-
Max Indexing Capacity (in OCUs) – Default is 96.
-
Min Search Capacity (in OCUs) – Optional. Leave blank for no minimum.
-
Max Search Capacity (in OCUs) – Default is 96.
-
Configure encryption
Choose an encryption option:
-
Use AWS owned key (default) – Amazon OpenSearch Service manages the encryption key at no additional cost.
-
Use customer managed key – Select a AWS KMS key from your account.
(Optional) Select the Customize encryption settings checkbox to choose or create a different AWS KMS key.
For more information about encryption policies, see Encryption in Amazon OpenSearch Serverless.
Configure network access
Configure network access for your collection:
-
Enable or disable public access to the OpenSearch endpoint.
-
(Optional) Add VPC endpoint rules to restrict access.
Configure data access policy
Choose how to configure data access:
-
Create new policy – Define a new data access policy:
-
Enter a policy name and optional description.
-
Use the visual editor to add statements with principals and resource permissions, or switch to the JSON editor for direct policy editing.
-
Add principals: IAM users and roles, SAML users and groups, or users and groups.
-
Specify resource permissions for collections and indexes.
-
-
Add to existing policy – Select an existing data access policy from the dropdown.
Configure additional settings
-
(Optional) Add tags to your collection as key-value pairs.
Configure OpenSearch UI settings
Configure the OpenSearch application and workspace for your collection:
-
For OpenSearch application selection, choose one of the following:
-
Select existing OpenSearch application – Choose an existing application from the OpenSearch application name dropdown. Then for Workspace selection, choose Select existing workspace or Create new workspace and enter a name.
-
Create new OpenSearch application – Enter a name for the new application. You must also create a new workspace – enter a name for the workspace.
-
To create the collection, review your configuration and choose Submit.
After you submit, the collection enters a Creating status. The console also creates the collection group (if new), encryption policy, network policy, and data access policy as needed. When the collection status changes to Active, it is ready to use. This typically takes several minutes.
Create a Classic collection
Classic collections use a multi-step wizard with per-collection security, network, and encryption configuration. Each collection is configured independently.
Configure collection settings (Classic)
To configure basic collection settings (Classic)
-
Navigate to the Amazon OpenSearch Service console at https://console.aws.amazon.com/aos/home/
. In the left navigation pane, expand Serverless and choose Collections. -
Choose Create collection.
The console opens the Create NextGen collection form by default.
-
In the Serverless generation field, choose Switch to Classic.
The console navigates to the Classic collection creation wizard. The Serverless generation field now displays Classic with a description: This is the Classic collections create flow.
Tip
To switch back to NextGen, choose Switch to NextGen in the Serverless generation field.
-
Provide a name and description for the collection. The name must meet the following criteria:
-
Is unique to your account and AWS Region
-
Contains only lowercase letters a–z, the numbers 0–9, and the hyphen (-)
-
Contains between 3 and 32 characters
-
-
Choose a collection type:
-
Time series – Log analytics segment that focuses on analyzing large volumes of semi-structured, machine-generated data. At least 24 hours of data is stored on hot indexes, and the rest remains in warm storage.
-
Search – Full-text search that powers applications in your internal networks and internet-facing applications. All search data is stored in hot storage to ensure fast query response times. Choose this option if you are enabling automatic semantic search.
-
Vector search – Semantic search on vector embeddings that simplifies the management of vector data. Powers machine learning (ML) augmented search experiences and generative AI applications such as chatbots, personal assistants, and fraud detection.
-
-
(Optional) Configure deployment options:
-
Standby replicas – Choose whether to enable standby replicas for high availability.
-
GPU acceleration – For vector search collections, choose whether to enable GPU acceleration.
-
-
Choose Next.
To configure security and access policies (Classic)
-
Encryption policy – Choose how to encrypt data in your collection:
-
Use AWS owned key (default) – Amazon OpenSearch Service manages the encryption key at no additional cost.
-
Use customer managed key – Select a AWS KMS key from your account for encryption.
(Optional) Select Use existing policy if a matching encryption policy already exists for your collection name.
-
-
Network access – Configure how users and applications access your collection:
-
Enable or disable access to OpenSearch endpoint.
-
Enable or disable access to OpenSearch Dashboards .
-
Choose Public access or restrict access through VPC endpoints.
-
-
Data access policy – Define which principals can access your collection's data:
-
Create new policy – Define a new data access policy using the visual editor or JSON editor. Add principals (IAM users, IAM roles, or SAML users and groups) and specify resource permissions.
-
Add to existing policy – Select an existing data access policy from the dropdown.
-
-
Choose Next.
Configure OpenSearch UI (Classic)
Configure the OpenSearch application and workspace for your collection:
-
For OpenSearch application selection, choose one of the following:
-
Select existing OpenSearch application – Choose an existing application from the OpenSearch application name dropdown. Then for Workspace selection, choose Select existing workspace or Create new workspace and enter a name.
-
Create new OpenSearch application – Enter a name for the new application. You must also create a new workspace – enter a name for the workspace.
-
Choose Next.
Configure additional search fields (Classic)
The options you see on this page depend on the type of collection you are creating.
-
Search collections – Configure automatic semantic enrichment and lexical search fields.
-
Time series collections – Configure time series search fields.
-
Vector search collections – Configure vector fields, including field name, dimensions, and distance metric.
(Optional) Configure the search fields relevant to your collection type. Choose Next. Review all settings on the summary page and choose Submit to create the collection.
After you submit, the collection enters a Creating status. When the status changes to Active, the collection is ready to use.
Topics
Configure automatic semantic enrichment
When you create or edit a collection, you can configure automatic semantic enrichment, which simplifies semantic search implementation and capabilities in Amazon OpenSearch Service. Semantic search returns query results that incorporate not just keyword matching, but the intent and contextual meaning of the user's search. For more information, see Automatic semantic enrichment for Serverless.
To configure automatic semantic enrichment
-
In the Index details section, for Index name, specify a name.
-
In the Automatic semantic enrichment fields section, choose Add semantic search field.
-
In the Input field name for semantic enrichment field, enter the name of a field that you want to enrich.
-
Data type is Text. You can't change this.
-
For Language, choose either English or Multilingual.
-
Choose Add field.
-
After you finish configuring optional fields for your collection, choose Next.
-
Review your changes and choose Submit to create the collection.
Configure time series search fields
The options in the Time series search fields section pertain to time series data and data streams. For more information about these subjects, see Managing time-series data in Amazon OpenSearch Service with data streams.
To configure time series search fields
-
In the Time series search fields section, choose Add time series field.
-
For Field name, enter a name.
-
For Data type, choose a type from the list.
-
Choose Add field.
-
After you finish configuring optional fields for your collection, choose Next.
-
Review your changes and choose Submit to create the collection.
Configure lexical search fields
Lexical search seeks an exact match between a search query and indexed terms or keywords.
To configure lexical search fields
-
In the Lexical search fields section, choose Add search field.
-
For Field name, enter a name.
-
For Data type, choose a type from the list.
-
Choose Add field.
-
After you finish configuring optional fields for your collection, choose Next.
-
Review your changes and choose Submit to create the collection.
Configure vector search fields
Note
The Engine property is only configurable with Classic collections
and is unsupported for NextGen collections.
To configure vector search fields
-
In the Vector fields section, choose Add vector field.
-
For Field name, enter a name.
-
For Engine, choose a type from the list.
-
Enter the number of dimensions.
-
For Distance Metric, choose a type from the list.
-
After you finish configuring optional fields for your collection, choose Next.
-
Review your changes and choose Submit to create the collection.