AWS wellarchitected documentation change
Summary
Updated sustainability best practices for data storage: 1) Improved formatting of headings, 2) Added generative AI storage optimization techniques, 3) Enhanced storage class guidance with S3 examples, 4) Fixed product name references, 5) Added data purification and columnar format recommendations
Security assessment
Changes focus solely on sustainability improvements (carbon footprint reduction) and storage optimization for generative AI workloads. No security vulnerabilities, incidents, or security controls are mentioned. The additions about data classification and tagging are for cost/energy efficiency, not security compliance.
Diff
diff --git a/wellarchitected/latest/financial-services-industry-lens/fsisus10.md b/wellarchitected/latest/financial-services-industry-lens/fsisus10.md index e3494680b..e34c225f6 100644 --- a//wellarchitected/latest/financial-services-industry-lens/fsisus10.md +++ b//wellarchitected/latest/financial-services-industry-lens/fsisus10.md @@ -5 +5 @@ -FSISUS10-BP01 Balance your data performance requirements against its carbon footprint FSISUS10-BP02 Separate data into hot, warm, cold storage +FSISUS10-BP01 Balance your data performance requirements against its carbon footprintFSISUS10-BP02 Separate data into hot, warm, and cold storage @@ -13 +13 @@ Data is at the heart of strategic innovations for the financial services industr -**Prescriptive guidance** +### Prescriptive guidance @@ -17 +17 @@ To balance data performance requirements against its carbon footprint: - * Define proxy metrics to monitor the business outcome of the data-involved service in relation to their environmental impact. An example proxy metric could be efficiency of the AI/ML service to help detect fraud faster (with the associated cost saving) and the carbon footprint of training and storing the data. These proxy metrics then become the vehicle to balance your performance requirements against its carbon footprint. Proxy metrics can be collected by importing AWS Cost and Usage Report as well as Amazon CloudWatch metrics into Amazon S3 and monitored using Amazon Athena and Amazon Quick Suite. + * Define proxy metrics to monitor the business outcome of the data-involved service in relation to their environmental impact. An example proxy metric could be efficiency of the AI/ML service to help detect fraud faster (with the associated cost saving) and the carbon footprint of training and storing the data. These proxy metrics then become the vehicle to balance your performance requirements against its carbon footprint. Proxy metrics can be collected by importing AWS Cost and Usage Report as well as Amazon CloudWatch metrics into Amazon S3 and monitored using Amazon Athena and Quick Suite. @@ -19 +19,6 @@ To balance data performance requirements against its carbon footprint: - * Use the right storage class for Amazon S3 Storage Classes based on the data performance requirements. The storage class impacts the environmental impact of the dataset through its access patterns and its architecture. For example, in [Amazon S3 One Zone-IA](https://docs.aws.amazon.com/AmazonS3/latest/userguide/storage-class-intro.html), energy and server capacity are reduced because data is stored only within one Availability Zone. Amazon S3 Storage Classes can be configured at the object level and a single bucket can contain objects stored across all of the storage classes. + * Use the right storage class for Amazon S3 Storage Classes based on the data performance requirements. The storage class impacts the environmental impact of the dataset through its access patterns and its architecture. For example, in [Amazon S3 One Zone-IA](https://docs.aws.amazon.com/AmazonS3/latest/userguide/storage-class-intro.html), energy and server capacity are reduced because data is stored only within one Availability Zone. Amazon + + + + +S3 Storage Classes can be configured at the object level and a single bucket can contain objects stored across all of the storage classes. @@ -23 +28 @@ To balance data performance requirements against its carbon footprint: - * You can also use Amazon S3 lifecycle policies to transition objects automatically between storage classes without application changes. In general, you have to make a trade-off between resource efficiency, access latency, and reliability when considering these storage mechanisms. + * You can also use Amazon S3 lifecycle policies to transition objects automatically between storage classes without application changes. In general, you must make a trade-off between resource efficiency, access latency, and reliability when considering these storage mechanisms. @@ -43 +48 @@ Data types may include the following: -## FSISUS10-BP02 Separate data into hot, warm, cold storage +## FSISUS10-BP02 Separate data into hot, warm, and cold storage @@ -45 +50 @@ Data types may include the following: -**Prescriptive guidance** +### Prescriptive guidance @@ -61 +66,7 @@ Data types may include the following: - * [AWS Glue Data Catalog](https://docs.aws.amazon.com/glue/latest/dg/components-overview.html#data-catalog-intro) lets you store, annotate, and share metadata in the AWS cloud while providing comprehensive audit and governance capabilities, in order to periodically audit your environment for untagged and unclassified data and tag the data appropriately. + * [AWSAWS Glue Data Catalog](https://docs.aws.amazon.com/glue/latest/dg/components-overview.html#data-catalog-intro) lets you store, annotate, and share metadata in the AWS cloud while providing comprehensive audit and governance capabilities, to periodically audit your environment for untagged and unclassified data and tag the data appropriately. + + * Optimize storage for generative AI training data and model artifacts using appropriate storage classes. + + * Implement data purification filters to reduce unnecessary generative AI training data storage. + + * Use columnar formats and compression for generative AI datasets.