Cloud Pro: AWS Containers: Choose, Compare and Launch

What's inside

1. Level 1: Every AWS Container Service on One PageFree
   Lay out all of AWS's container services on one map by category (Kubernetes, serverless containers, single containers, PaaS, OpenShift, registry, security), with their current and former names and their lifecycle status as of October 2026: App Runner closed to new customers on 30 April 2026, and App Mesh support ended on 30 September 2026.
2. Level 2: Building a Standard Container Workload ProfileFree
   Write a reusable workload profile that records number of services, state, traffic shape and scaling, GPUs, team Kubernetes skills and the operating model it can keep up, portability, compliance and cost model, so every later choice is checked against the same facts.
3. Level 3: Reading Dated Quotas, Limits and Pricing Pages
   Learn to collect evidence the way the rest of the course does: find the authoritative quotas and pricing pages, tell adjustable quotas from hard limits, and record every number with its source, Region and date.
4. Level 4: Amazon EKS: The Managed Control Plane
   Understand what Amazon EKS actually manages and what it leaves to you: the control plane, Kubernetes version support, standard versus extended support, access entries and add-ons. Also see where EKS fits a workload and where it does not.
5. Level 5: EKS Limits, Pricing and Cost Levers
   Work through the EKS limits that cause problems in production and its cost structure: the hourly cluster fee, the extra fee for extended support, the cost of nodes, and the levers that reduce the bill. Use dated examples from the docs.
6. Level 6: EKS Quick Start with Managed Node Groups
   Launch the smallest working EKS cluster with a managed node group using eksctl and the AWS CLI, then again with Terraform or CloudFormation. Deploy a sample app, tear everything down cleanly, and finish with a pros and cons table for EKS.
7. Level 7: Karpenter: Just-in-Time Nodes on EKS
   Learn how Karpenter provisions and consolidates EC2 capacity straight from pending pods using NodePools and EC2NodeClasses, and when it beats managed node groups and Cluster Autoscaler.
8. Level 8: EKS Auto Mode: AWS Runs the Data Plane
   Learn how EKS Auto Mode gives AWS the job of running nodes, core add-ons, load balancing and storage. Cover what Auto Mode adds to the bill, what you can no longer customise, and when it is worth the extra fee.
9. Level 9: EKS on Fargate, EKS Anywhere and Hybrid Nodes
   Cover the less common EKS options in brief: Fargate profiles for serverless pods, EKS Anywhere for clusters on your own infrastructure, and EKS Hybrid Nodes for on-premises capacity attached to a cloud control plane, including each option's limits and when it is worth using.
10. Level 10: ROSA: Red Hat OpenShift Service on AWS
    Go through ROSA in nine parts: hosted control planes, the jointly supported operating model, pricing (OpenShift service fees on top of EC2), limits, quick start and clean-up, and when OpenShift is worth paying for instead of plain EKS.
11. Level 11: Amazon ECS: Clusters, Tasks, Services on EC2
    Learn the Amazon ECS model (task definitions, services, clusters, capacity providers) and run it on EC2 with Auto Scaling group capacity providers. Cover ECS's limits, why ECS has no control-plane fee, and what you still have to run yourself on EC2.
12. Level 12: AWS Fargate: Serverless Compute for Containers
    Study AWS Fargate as a compute engine in its own right: billing per vCPU and per GB of memory per second, Fargate Spot, Graviton, the allowed task sizes, ephemeral storage, platform versions, and the limits that rule it out for some workloads.
13. Level 13: ECS on Fargate Quick Start and Verdict
    Deploy the smallest working ECS service on Fargate behind an Application Load Balancer using the AWS CLI and then CloudFormation or Terraform, tear it down, and write the full ECS pros and cons table and verdict.
14. Level 14: ECS Service Connect after App Mesh
    Learn ECS Service Connect for service discovery and service-to-service traffic, and how to move away from App Mesh, whose support ended on 30 September 2026. Includes the blue/green recreate pattern, because a service cannot be in both a mesh and a Service Connect namespace.
15. Level 15: ECS Express Mode: The Simple Path on ECS
    Learn ECS Express Mode, which AWS recommends as the path away from App Runner. It takes a container image and creates a load-balanced, autoscaling Fargate service with sensible defaults that you can still customise with the rest of ECS.
16. Level 16: App Runner for Existing Customers
    Cover App Runner for teams that already use it: it closed to new customers on 30 April 2026 and gets no new features, but security and availability updates continue. Look at its pricing and limits, and at how to plan and carry out a move to ECS Express Mode.
17. Level 17: Lambda Container Images in Brief
    Learn when to package a Lambda function as a container image: the image size limit, the runtime interface, cold starts, request-based pricing, and the event-driven workloads where Lambda beats a long-running container.
18. Level 18: Amazon ECR: Registry, Replication and Cost
    Run Amazon ECR properly in production: private and public repositories, lifecycle policies, cross-Region and cross-account replication, pull-through cache, immutable tags, storage and data-transfer pricing, and the API rate limits that slow down large deployments.
19. Level 19: GuardDuty Runtime Monitoring for Containers
    Learn how Amazon GuardDuty protects running containers through EKS audit-log monitoring and Runtime Monitoring for EKS, ECS on Fargate and EC2: how the agent is deployed, what it detects, how it is priced, and how to triage findings.
20. Level 20: Amazon Inspector for Container Images
    Use Amazon Inspector to scan container images in ECR for vulnerabilities, both when pushed and continuously, including how scan duration is set, how findings are mapped to running workloads, and how it is priced. Then combine it with GuardDuty into one container security baseline.
21. Level 21: GuardDuty and Inspector for Containers
    Learn how Amazon GuardDuty (EKS audit log monitoring and Runtime Monitoring for EKS, ECS and Fargate) and Amazon Inspector (ECR image scanning) cover container threats and vulnerabilities, using the same nine-part pattern, a quick start and clean-up.
22. Level 22: Cluster and Account Topology
    Decide how many clusters and accounts to run and how to split tenants, environments and teams across them, weighing blast radius against operating cost.
23. Level 23: Capacity Design: Instances, Graviton, Spot, Fargate
    Design the compute underneath your containers by choosing instance families, Arm versus x86, the mix of Spot and On-Demand, and where Fargate fits instead of nodes.
24. Level 24: Autoscaling Pods, Tasks and Capacity Together
    Tune the two layers of scaling in production, the workload layer (HPA, KEDA, ECS service auto scaling) and the capacity layer (Karpenter, Auto Mode, capacity providers), so that they work together without fighting.
25. Level 25: Version Upgrades and Platform Lifecycle
    Run a repeatable upgrade process for EKS versions, add-ons and nodes, and for ECS agents and Fargate platform versions, including what extended support costs.
26. Level 26: VPC Networking and IP Address Planning
    Design VPCs and subnets for containers by understanding how the VPC CNI and awsvpc mode use IP addresses, and how to avoid running out of them.
27. Level 27: Ingress and Service-to-Service Traffic
    Choose how traffic gets into your services and moves between them, using ALB, NLB, the AWS Load Balancer Controller, Gateway API, ECS Service Connect and VPC Lattice.
28. Level 28: Image Supply Chain: Build, Sign, Verify
    Secure the path from source code to running image with minimal base images, SBOMs, image signing, ECR controls and admission policies that block untrusted images.
29. Level 29: Runtime Security, Identity and Secrets
    Lock down running containers with least-privilege IAM (EKS Pod Identity, IRSA, ECS task roles), secrets management, Pod Security Standards and network policies.
30. Level 30: GitOps and Progressive Delivery
    Deliver changes safely with GitOps tools such as Argo CD and Flux on EKS and with native rolling, blue/green and canary deployments on ECS, including automatic rollback.
31. Level 31: Observability: Metrics, Logs, Traces, Cost
    Instrument container platforms with CloudWatch Container Insights, OpenTelemetry, managed Prometheus and Grafana, and log routing, while keeping the cost of telemetry under control.
32. Level 32: Worked Cost: Two Contrasting Workloads
    Price two contrasting workloads end to end on EKS, ECS on EC2, ECS on Fargate and a serverless option, using dated example prices, and see which levers change the answer.
33. Level 33: Building a Weighted Selection Scorecard
    Turn the Level 2 workload profile into a weighted scorecard whose criteria, weights and scoring rules can be defended to stakeholders.
34. Level 34: Scoring the AWS Options and Testing Sensitivity
    Apply the scorecard to EKS (with node groups, Karpenter or Auto Mode), ECS on EC2, ECS on Fargate, ECS Express Mode, Lambda and ROSA, then check how robust the winner is.
35. Level 35: Scoring AWS Against Another Cloud
    Compare the chosen AWS option with the matching services on another cloud (for example GKE and Cloud Run, or AKS and Azure Container Apps) using the same scorecard and dated evidence.
36. Level 36: Portability and Exit Planning
    Plan for portability and exit by finding the lock-in points in each option and deciding which ones are worth accepting.
37. Level 37: Final Challenge: Decision Workshop Exam
    Bring the whole course together in a decision workshop: profile a new workload, filter and score the AWS options and another cloud, cost the shortlist, design it for production and record the decision on one page.

Access

The first 2 levels are free with a free account. Every level, the podcast edition and the AI tutor come with All Access at £4.99/month or any Creator plan — see pricing.

Start free →