AWS solutions documentation change
Summary
Updated migration instructions for v2.0.0+ versions, clarified version requirements, and modified references to deprecated integrations
Security assessment
Changes focus on version compatibility and migration procedures rather than security features. While deprecating third-party integrations (Anthropic/Hugging Face) in favor of AWS services (Bedrock/SageMaker) could have security implications, the documentation update does not explicitly mention security motivations.
Diff
diff --git a/solutions/latest/generative-ai-application-builder-on-aws/update-the-solution.md b/solutions/latest/generative-ai-application-builder-on-aws/update-the-solution.md index f31c208ae..dda61deb1 100644 --- a//solutions/latest/generative-ai-application-builder-on-aws/update-the-solution.md +++ b//solutions/latest/generative-ai-application-builder-on-aws/update-the-solution.md @@ -5 +5 @@ -Step 1: Update Deployment dashboardStep 2: Migrate use case configurationsStep 3: Update use cases +Step 1: Update Deployment dashboardStep 2: Migrate use case configurations (Only updates from versions below 2.0.0)Step 3: Update use cases @@ -22 +22 @@ If you have previously deployed the solution, follow this procedure to update th - 1. In v2.0.0, we are deprecating direct integration with Anthropic and Hugging Face, in favor of Amazon Bedrock (Anthropic models hosted through Bedrock) and Amazon SageMaker. You can deploy models available through Hugging Face through SageMaker JumpStart. Refer to [Use Hugging Face with Amazon SageMaker](https://docs.aws.amazon.com/sagemaker/latest/dg/hugging-face.html) for more details. + 1. In v2.0.0, integration with Anthropic and Hugging Face was deprecated in favor of Amazon Bedrock and Amazon SageMaker AI. You can deploy models available through Hugging Face through SageMaker JumpStart. Refer to [Use Hugging Face with Amazon SageMaker AI](https://docs.aws.amazon.com/sagemaker/latest/dg/hugging-face.html) for more details. @@ -62 +62 @@ You can view the status of the stack in the AWS CloudFormation console in the ** -This update will create a web UI stack (which replaces `amplify-ui` implementation of the login screen with Cognito hosted UI) and new CloudFront URL, which can be obtained from the Output section of the CloudFormation console once the stack status is UPDATE_COMPLETE. +If the existing Solution version was prior to v2.0.0, updating will create a web UI stack (which replaces the `amplify-ui` implementation of the login screen with a Cognito hosted UI) and a new CloudFront URL, which can be obtained from the Output section of the CloudFormation console once the stack status is UPDATE_COMPLETE. @@ -66 +66 @@ This update will create a web UI stack (which replaces `amplify-ui` implementati -Existing use cases created using versions prior to v2.0.0 will NOT be displayed until you completed the steps outlined below. +Existing use cases created using versions prior to v2.0.0 will NOT be displayed until you complete the steps outlined below. @@ -68 +68 @@ Existing use cases created using versions prior to v2.0.0 will NOT be displayed -## Step 2: Migrate use case configurations +## Step 2: Migrate use case configurations (Only updates from versions below 2.0.0) @@ -70 +70 @@ Existing use cases created using versions prior to v2.0.0 will NOT be displayed -The schema for storing and the AWS service to store use case configuration has changed in this version. Follow the steps described in [GAAB v2 Migration User Guide](https://github.com/aws-solutions/generative-ai-application-builder-on-aws/tree/main/docs/migration/migration.md) using the [gaab_v2_migration.py](https://github.com/aws-solutions/generative-ai-application-builder-on-aws/blob/main/source/scripts/v2_migration/gaab_v2_migration.py) script. After you run the script, you can access the Deployment dashboard to view the deployed use cases. +The schema for storing and the AWS service to store use case configuration has changed in version 2.0.0. Follow the steps described in [GAAB v2 Migration User Guide](https://github.com/aws-solutions/generative-ai-application-builder-on-aws/tree/main/docs/migration/migration.md) using the [gaab_v2_migration.py](https://github.com/aws-solutions/generative-ai-application-builder-on-aws/blob/main/source/scripts/v2_migration/gaab_v2_migration.py) script. After you run the script, you can access the Deployment dashboard to view the deployed use cases. @@ -78 +78 @@ You must follow the steps below to complete migrating the use cases. -You can edit the deployed use cases with new features available in v2.0.0. See [Use the solution](https://docs.aws.amazon.com/solutions/latest/generative-ai-application-builder-on-aws/use-the-solution.html) for information about how to use the features in this solution. +You can edit the deployed use cases with new features available in the latest versions of GAAB. See [Use the solution](https://docs.aws.amazon.com/solutions/latest/generative-ai-application-builder-on-aws/use-the-solution.html) for information about how to use the features in this solution. @@ -80 +80 @@ You can edit the deployed use cases with new features available in v2.0.0. See [ -To update use cases to v2.0.0, you must complete the `Edit`use case steps in the Deployment dashboard (although you might not make any changes). This action triggers a CloudFormation stack update and uses the latest template version, which will upgrade the use cases to v2.0.0. +To update use cases to the latest version, you must complete the `Edit`use case steps in the Deployment dashboard (although you might not make any changes). This action triggers a CloudFormation stack update with the latest template version. @@ -84 +84 @@ To update use cases to v2.0.0, you must complete the `Edit`use case steps in the -Use cases created with 1.x versions of the solution might not work with later versions. Hence, we recommend cloning existing use cases created with versions prior to v2.0.0 through the Deployment dashboard. Then, gradually migrate and replace with new use cases created using v2.0.0 or later. +Use cases created with 1.x or 2.x versions of the solution might not work with later versions. Hence, we recommend cloning existing use cases created with versions prior to v3.0.0 through the Deployment dashboard. Then, gradually migrate and replace with new use cases created using v3.0.0 or later.