Notes on cloud migration, Google Cloud Platform, and running infrastructure that doesn't fight back.
How we helped a client move off a monolithic LAMP stack and onto autoscaling microservices on GKE, externalizing data services along the way, to scale out as demand grew from a handful of customers to thousands.
Read more →
ingress-nginx is being retired. Here's why that's the real trigger for moving clients to GKE's Gateway API, how we map Ingress and its annotations onto HTTPRoute and GCPBackendPolicy, and how consolidating to a wildcard certificate per Gateway (with Vault PKI for internal traffic) simplifies every application's deployment.
Read more →
How we decouple application deployments from Terraform-driven infrastructure changes with a GitOps workflow using Flux — from bootstrapping GitRepository and Kustomization per tenant to letting helm-controller own HelmRelease lifecycles — backed by GitLab's Kubernetes Agent Server for secure access, in-GitLab operational visibility, and commit-status validation of every deploy.
Read more →