umac-ai-deploy-prod

Deploying full-stack applications to AWS ECS and S3 requires a standardized procedure covering Docker image builds, ECS service updates, S3 static asset synchronization, and CloudFront cache invalidation. This ensures consistent backend and frontend production updates without manual deployment errors.

ECS task crashes after production deployment often stem from misconfigured container environments or failed health checks. Troubleshooting involves verifying Docker image integrity, checking ECS task definitions, and reviewing application logs to identify the exact startup failure.

Fixing CloudFront 502 errors and TLS misconfigurations requires verifying origin server health and ensuring SSL certificates are correctly attached to your distribution. Invalidating the CloudFront cache after updating frontend S3 static assets is also required to resolve delivery failures.

The correct order to run Prisma database migrations during AWS CDK deployments is after infrastructure provisioning but before updating ECS services. This sequence guarantees the database schema is updated before new backend containers attempt to access the modified tables.

Detecting AWS infrastructure drift before deploying updates requires running a quick probe check across core services to identify if resources have migrated off AWS. This prevents wasted effort on outdated infrastructure configurations and ensures targeting the active environment.

SES email delivery failures in production environments usually occur due to unverified domain identities or exceeded sending limits. Checking your AWS SES configuration and verifying sender domain policies will resolve common production email routing issues.