Cloud Pro: Google Cloud Containers: Choose, Compare and Launch

What's inside

1. Level 1: Every Google Cloud Container Service on One PageFree
   Learn the full Google Cloud container landscape: Kubernetes, serverless containers, single containers on VMs, PaaS, OpenShift, registries and security. You also learn the current and former names, so you can read old docs, old Terraform and job adverts without confusion.
2. Level 2: The Standard Container Workload ProfileFree
   Build the reusable workload profile that every later decision in the course depends on: number of services, stateless or stateful, traffic shape and scaling, GPUs, Kubernetes skills and the operating model your team can keep up, portability, compliance and cost model.
3. Level 3: GKE Standard: Full Control, Your Nodes
   Go through GKE Standard in nine parts: what it is, where it fits, the limits that bite, the cluster management fee and node pricing, what you operate and what Google operates, pros and cons, real uses, a quick start with gcloud, kubectl and Terraform, and when to choose it or avoid it.
4. Level 4: GKE Autopilot: Pay for Pods, Not Nodes
   Go through GKE Autopilot in nine parts, focusing on Pod-based billing, compute classes, the workload restrictions it enforces, and how its operating model differs from GKE Standard.
5. Level 5: Fleets and the Former GKE Enterprise Features
   Learn how fleets group clusters, and where the former GKE Enterprise features went after Google merged them into the single GKE offering in September 2025: Config Sync, Policy Controller, team management and multi-cluster networking.
6. Level 6: Cloud Run Services: Request-Driven Containers
   Go through Cloud Run services in nine parts: request-based versus instance-based billing, concurrency, cold starts, revisions and traffic splitting, the limits that bite, and the smallest deployment with gcloud and Terraform.
7. Level 7: Cloud Run Jobs: Run-to-Completion Containers
   Learn when to use Cloud Run jobs for batch work: tasks and parallelism, retries, task timeouts, scheduling with Cloud Scheduler, and how jobs compare with Kubernetes Jobs on GKE.
8. Level 8: Cloud Run Worker Pools: Pull-Based Background Work
   Learn Cloud Run worker pools, which became generally available in April 2026. They are for always-on, pull-based consumers such as Pub/Sub pull, Kafka or RabbitMQ workers: they have no public endpoint, scale manually and are billed per instance.
9. Level 9: GPUs on Cloud Run: Serverless Inference
   Learn how to attach GPUs to Cloud Run: which GPU types and regions are supported, the minimum CPU and memory each needs, cold-start and model-loading trade-offs, quota, and how the cost compares with GPU node pools on GKE.
10. Level 10: Artifact Registry: Where Every Image Lives
    Go through Artifact Registry in nine parts: repository formats and modes (standard, remote and virtual), vulnerability scanning, cleanup policies, IAM, storage and egress pricing, and migrating off the shut-down Container Registry, including gcr.io domain support.
11. Level 11: Cloud Service Mesh: Traffic, Identity, Telemetry
    Go through Cloud Service Mesh in nine parts: the managed control plane, sidecar versus proxyless data planes, mTLS and service identity, traffic policies, supported platforms, pricing, and when a mesh is worth its operational cost.
12. Level 12: Binary Authorization: Only Trusted Images Run
    Learn Binary Authorization for GKE and Cloud Run: policies, attestors and attestations, continuous validation, break-glass deployment, and the operational pitfalls of enforcing deploy-time trust.
13. Level 13: Cloud Build and Cloud Deploy in Brief
    Learn, at the level a service selection needs, how Cloud Build builds images and Cloud Deploy promotes releases to GKE and Cloud Run: build configs, private pools, build provenance, delivery pipelines, targets, approvals and pricing.
14. Level 14: GKE on Other Clouds and On-Premises in Brief
    Learn the current options for running GKE-managed Kubernetes outside Google Cloud. These are GKE attached clusters and Google Distributed Cloud (software only, for VMware and bare metal). You also learn the dated retirement of GKE on AWS and GKE on Azure: in maintenance mode since 17 March 2025, with shutdown on 17 March 2027.
15. Level 15: Container Threat Detection and the Retirement Register
    Learn what Security Command Center Container Threat Detection catches at runtime on GKE, which service tiers include it and how its findings flow, then build a dated register of the container services and features that are retiring or have retired.
16. Level 16: Cluster and Node Pool Design for Production
    Design production GKE clusters the way experience teaches: regional versus zonal control planes, how many clusters, multi-tenancy by namespace or by cluster, node pool layout by workload class, machine families, and Spot capacity.
17. Level 17: Autoscaling That Behaves Under Real Load
    Learn how the autoscaling layers work together and how they fail: Horizontal and Vertical Pod Autoscaling, cluster autoscaler, node auto-provisioning and compute classes on GKE, and concurrency and instance limits on Cloud Run.
18. Level 18: Upgrades, Release Channels and Maintenance
    Run GKE upgrades without drama: release channels, maintenance windows and exclusions, surge versus blue-green node upgrades, API deprecation checks, and how Cloud Run revisions avoid the problem entirely.
19. Level 19: Container Networking: VPC-Native, IPs and Policy
    Learn container networking on Google Cloud from the ground up: VPC-native clusters and IP address planning, GKE Dataplane V2 and network policy, private clusters, Shared VPC, and Cloud Run egress through Direct VPC egress or connectors.
20. Level 20: Ingress and Load Balancing with Gateway API
    Expose container workloads properly: GKE Gateway API versus Ingress, external and internal Application Load Balancers, multi-cluster gateways, Cloud Run ingress settings, serverless network endpoint groups (NEGs), Cloud Armor and Identity-Aware Proxy (IAP).
21. Level 21: Image Supply Chain: From Commit to Signed Image
    Build a trusted path for container images, from source to a scanned, signed image in Artifact Registry that Binary Authorization admits, and see where each control catches a bad image.
22. Level 22: Runtime Security and Workload Hardening
    Lock down running containers with least-privilege identity, pod security, network policy, sandboxing and threat detection, and know who responds when an alert fires.
23. Level 23: GitOps: Declarative Config Across Clusters
    Run cluster and workload configuration from Git with Config Sync or Argo CD, organise repos for a fleet, and handle drift, secrets and emergency changes.
24. Level 24: Progressive Delivery and Safe Rollouts
    Ship changes in steps with Cloud Deploy pipelines, canary and blue-green rollouts on GKE and Cloud Run, and automated rollback based on real signals.
25. Level 25: Observability: Metrics, Logs, Traces and SLOs
    Instrument container workloads with Cloud Logging, Managed Service for Prometheus and OpenTelemetry traces, then turn the signals into SLOs and alerts people trust.
26. Level 26: Worked Cost 1: A Stateless Web API
    Cost a request-driven web API on Cloud Run, GKE Autopilot and GKE Standard with dated example prices, showing where each option wins as traffic grows.
27. Level 27: Worked Cost 2: A Stateful and GPU Workload
    Cost a mixed platform of stateful services and GPU inference or batch jobs on GKE Standard, Autopilot and Cloud Run GPUs, including storage, accelerators and operating effort.
28. Level 28: The Weighted Selection Scorecard
    Turn the workload profile from Level 2 into weighted criteria and score each Google Cloud container option against them in a repeatable, defensible way.
29. Level 29: Gathering Dated Cross-Cloud Evidence
    Collect comparable, dated evidence on another cloud's container services (managed Kubernetes, serverless containers, registry, mesh) from primary sources so the comparison is fair.
30. Level 30: Scoring Google Cloud Against Another Cloud
    Apply the weighted scorecard to Google Cloud and the other cloud side by side, adding portability, skills and exit costs, and reach a recommendation you can defend.
31. Level 31: Final Challenge: The Decision Workshop Exam
    Take a new scenario from workload profile through service choice, production design, cost and cross-cloud scoring to a one-page container service selection record you can defend under questioning.

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 →