Our team had a rock-solid, well-engineered microservices platform on AWS—complete with Kubernetes, GitLab CI/CD, Terraform, Kafka, Postgres, Prometheus, and even a lightweight LinkerD mesh—so we figured migrating to GCP would be a breeze. Spoiler: it wasn’t! From subtle Kubernetes quirks to data-replication hiccups and reconfiguring Prometheus and Fluentd, every step demanded painstaking re-engineering.
Dive into our “oops” moments and clever hacks as we retune Terraform modules, twist cluster settings, and finally flip that DNS record to GCP—proving that even the cleanest, containerized architecture can hit unexpected snags in a new cloud.
Watch on YouTube
Top comments (0)