Decoupling services and processing data at scale are what separate a fragile monolith from a resilient distributed system. This pillar covers the messaging backbone (SQS, SNS, Kinesis) that lets services fail independently without losing data, the analytics services that turn raw data into queryable insight, and the two ways AWS runs containers in production.
What This Pillar Covers
- SQS and SNS — decoupling producers from consumers, fan-out patterns, dead-letter queues
- Kinesis Data Streams and Firehose — real-time event ingestion and delivery to storage/analytics
- Athena and Redshift — serverless SQL over S3 data lakes vs. a managed data warehouse
- ECS — task definitions, service auto-scaling, and capacity providers for container workloads
- EKS and Fargate — managed Kubernetes control planes and serverless container compute
Who This Is For
Backend and platform engineers building event-driven architectures, and teams choosing between ECS, EKS, and Fargate for running containerized workloads on AWS.