Port the Whole Platform to Azure
Same container, same shape, different cloud. Here is what does not translate.
# what you build
One codebase, second cloud. You will rebuild the platform on Azure — AKS, Container Apps, Azure Database for PostgreSQL and Front Door — from the same container image and the same Terraform structure, and keep an honest running list of everything that did not port cleanly.
What you end up with
handles after this
~50,000 people at once
what it costs to run
measured against AWS
What you will be able to do
- Map every AWS piece you built to its Azure equivalent, and spot the gaps
- Terraform modules structured so a second cloud is a new file, not a rewrite
- Identity and secrets on Azure, without pasting credentials anywhere
- GPU availability and quota on Azure — the part nobody warns you about
- What the identical workload costs here versus AWS, at the same load
The build, step by step
These are the chapters of the video, in order.
- 1Stand up the clusterAKS with the same node pools and the same Helm chart.
- 2Move the dataAzure Database for PostgreSQL with pgvector, and the migration path.
- 3Put Front Door in frontRouting, TLS and the streaming settings, again — differently.
- 4Write down what brokeThe honest list of gaps, quotas and surprises, for your notes and mine.
The stack
- AKS
- Front Door
- Azure Postgres
- Terraform
# before this one
This project continues the system built in 07 · Observability & Cost Control for AI. You can start here, but the repo assumes the previous rung exists.
$ subscribe --new-articles
Get new articles in your inbox
One email when a new hands-on guide goes live — Kubernetes, AWS, CI/CD, MLOps. No spam, unsubscribe anytime.