Migrate from AWS

Overview

AWS gives you infinite flexibility. It also gives you infinite complexity. Moving to RaidFrame means keeping the production-grade infrastructure while eliminating the IAM policies, VPC configurations, CloudFormation templates, and $500/mo DevOps engineer time.

What Replaces What

Step 1: Install the CLI

curl -fsSL https://get.raidframe.com | sh
rf auth login

Step 2: Export from AWS

ECS Task Definition → raidframe.yaml

If you're running ECS, your task definition maps directly:

// AWS ECS task definition (simplified)
{
  "containerDefinitions": [{
    "name": "api",
    "image": "123456.dkr.ecr.us-east-1.amazonaws.com/api:latest",
    "cpu": 512,
    "memory": 1024,
    "portMappings": [{"containerPort": 8080}],
    "environment": [
      {"name": "DATABASE_URL", "value": "postgresql://..."},
      {"name": "REDIS_URL", "value": "redis://..."}
    ]
  }]
}

Becomes:

# raidframe.yaml
services:
  api:
    type: web
    build:
      dockerfile: Dockerfile
    port: 8080
    resources:
      cpu: 1
      memory: 1GB
    scaling:
      min: 2
      max: 10

RDS → RaidFrame Postgres

# Export from RDS
pg_dump -h your-rds-endpoint.amazonaws.com -U postgres -d mydb > backup.sql

# Import to RaidFrame
rf add postgres --plan pro
rf db import main backup.sql

ElastiCache → RaidFrame Redis

rf add redis --plan standard
# Redis data is ephemeral/cache — repopulate from your app on first requests

S3 → RaidFrame Storage

rf add storage

# Sync from S3
aws s3 sync s3://my-bucket ./local-copy/
rf storage sync ./local-copy/ s3://my-app-uploads/

Lambda → RaidFrame Services

# Lambda cron → RaidFrame cron
services:
  nightly-cleanup:
    type: cron
    schedule: "0 3 * * *"
    command: node cleanup.js
    timeout: 300s

Environment Variables

# Export from AWS Systems Manager Parameter Store
aws ssm get-parameters-by-path --path /myapp/prod --with-decryption \
  --query "Parameters[*].[Name,Value]" --output text | \
  awk '{split($1,a,"/"); printf "%s=%s\n", a[length(a)], $2}' > .env

# Import to RaidFrame
rf env push

Step 3: Deploy

rf deploy

What You Stop Paying For

Typical AWS bill for a small SaaS vs RaidFrame:

And that's before you count the 20+ hours/month managing it all.

What You Stop Managing

• IAM roles and policies
• VPC, subnets, security groups, NAT gateways
• ECS task definitions, services, clusters
• Auto-scaling policies and CloudWatch alarms
• ALB target groups, listeners, health checks
• RDS parameter groups, snapshots, maintenance windows
• CloudFormation / Terraform state
• ECR image lifecycle policies
• CloudWatch log groups, metric filters, dashboards
• Secrets Manager rotation schedules
• Certificate Manager validation

All of this is handled automatically on RaidFrame.

Common Migration Issues

VPC Networking

AWS apps often rely on VPC private networking. On RaidFrame, all services in a project share a private network automatically. Replace AWS VPC endpoints with service.internal hostnames.

IAM-Based Auth

If your services authenticate with other AWS services via IAM roles, you'll need to switch to API keys or credentials stored in rf secrets set.

Multi-Region

rf regions add us-west-1
rf regions add eu-west-1

Simpler than managing multiple ECS clusters, RDS read replicas, and Route 53 failover policies across regions.