Down for less than four minutes a month: how AWS deploys code

Amazon Web Services (AWS) emphasizes automation and strategic deployment procedures to maintain its impressive service-level agreements (SLAs), which promise high uptime percentages, such as 99.9% for most services and up to 99.999% for others like DynamoDB Global Tables. Their automated incident management and deployment strategies include automated rollbacks, progressive rollouts, and slow rollouts (baking), which together minimize the potential impact of service changes and ensure high availability. These processes involve monitoring essential metrics and dependencies, implementing a "one-box" stage, and conducting deployments in waves to isolate potential failures. AWS also uses deployment blockers to prevent changes during critical times or active incidents, and pipelines for non-code changes like feature flags. Inspired by AWS, the startup Graphite has adopted some of these practices, such as extending pre-production bake times and timing deployments outside core business hours, while acknowledging the challenges of implementing AWS's scale-intensive processes.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.