Hybrid Cloud on Cloud Service Mesh & GKE

End-to-end infrastructure ownership for a mission-critical hybrid cloud environment powering all APIs of a major retail operation. The platform ran a fleet of GKE clusters — some on-premises (VMware), some on GCP — with Apigee handling API management across the entire estate.

The primary engineering challenge was upgrading the production GKE clusters from Kubernetes 1.4 all the way to version 1.14 with zero downtime for the retail APIs. To achieve this safely, a full Disaster Recovery (DR) replica was provisioned in parallel. Traffic was progressively migrated to the DR environment, validated under production load, and then fully cut over — eliminating the risk of a 10-version in-place upgrade on live systems.

As part of the modernization, legacy Anthos networking was replaced with Cloud Service Mesh, delivering improved observability, mTLS between services, and a consistent traffic management model across the hybrid environment. Terraform and Ansible automated the entire provisioning and configuration lifecycle.