Serverless architecture on AWS: Our experiences

Stanley Black and Decker's Digital Accelerator has spent the last year migrating existing applications, and creating new ones, using serverless architecture on AWS. Christopher Phillips explores the pros and cons of going serverless, as well as the tools and patterns you need and the caveats learned along the way.

Stanley Black and Decker’s Digital Accelerator has spent the last year migrating existing applications, and creating new ones, using serverless architecture on AWS. Christopher Phillips explores the pros and cons of going serverless, as well as the tools and patterns you need and the caveats learned along the way. Christopher covers a few different products the Stanley Black and Decker digital accelerator has released and the underlying architectural decisions behind them. All but one of these products are entirely serverless (using managed infrastructure rather than any server, VM, etc., that you have to manage), and the remaining one is being moved in that direction. Christopher also explains setting up REST endpoints via Lambda and the API gateway, messaging patterns using AWS IoT or SNS and SQS, authentication, identity, and resource access via Cognito and STS, and VPCs and proper security considerations and touches on more popular services such as file and data handling via S3, RDS, Dynamo, and Elasticache, along with their various trade-offs—for each, detailing use cases and patterns for proper access control, scaling, and availability concerns.