They’d built a dream platform on AWS—microservices with Kubernetes, Terraform-managed infrastructure, GitLab CI/CD, Kafka, Postgres, Linkerd, Prometheus and all the right automation. When it came time to move that setup to GCP, what looked like a no-brainer quickly turned into a series of “oops” moments.
From unexpected Kubernetes quirks and data-replication headaches to tweaking monitoring/logging pipelines, they learned that not everything works the same across clouds. After plenty of creative workarounds and lessons learned, they finally flipped the DNS and proved that a painstaking re-engineering can pay off—if you’re ready for the challenge.
Watch on YouTube
Top comments (0)