Inspiration
Eliminating Multi-Cloud Operational Friction: Provisioning enterprise-grade API gateways across complex multi-environment lifecycles (Dev, QA, UAT, Prod) usually suffers from configuration drift, manual deployment bottlenecks, and high infrastructure overhead.
Hybrid Enterprise Architecture: The goal was to build a standardized, repeatable blueprint combining Apigee Hybrid with Azure Kubernetes Service (AKS), enabling unified API management across cloud boundaries while leveraging Terraform and Azure DevOps for total automation.
What it does
Multi-Environment Infrastructure Automation: Automates the lifecycle of Apigee Hybrid on Azure AKS across Dev, QA, UAT, and Prod environments with environment-tailored High Availability (HA) topologies.
Guaranteed High Availability & State Resilience: Enforces multi-AZ worker node distribution and a 3-node Cassandra quorum state plane to maintain a 99.99% target SLA and prevent split-brain conditions.
Automated Progressive Promotion Pipeline: Uses a multi-stage Azure DevOps pipeline (azure-pipelines.yml) to automatically build, lint, and deploy API proxies across Dev and QA, while applying automated approval gates for UAT and Production.
How we built it
Infrastructure as Code (IaC): Structured modular Terraform scripts to provision Azure AKS clusters, dedicated Apigee runtime/data node pools, and network ingress components.
Kubernetes Orchestration: Configured Azure AKS with multi-Availability Zone distribution, Horizontal Pod Autoscalers (HPA), and specialized storage classes for Cassandra persistence.
CI/CD Automation: Built a 5-stage Azure DevOps pipeline (azure-pipelines.yml) integrating Node.js tools, apigeelint static code analysis, artifact packaging, and automated promotion across environments.
Challenges we ran into
Multi-AZ Cassandra Quorum Alignment: Ensuring Cassandra stateful sets across AKS nodes properly maintained quorum and latency bounds during node failovers.
Pipeline Approval & Secret Management: Securely mapping GCP service account credentials and environment secrets within Azure DevOps pipeline variables while enforcing strict manual approval gates for UAT and Prod environments.
Environment-Specific Resource Balancing: Optimizing node sizes and cost constraints for lower environments (Dev/QA) without compromising Production-level parity required for testing.
Accomplishments that we're proud of
Zero-Downtime Deployment Guarantee: Achieved seamless API proxy revisions and zero-downtime traffic shifts during Production deployments.
End-to-End Enterprise Automation: Successfully unified Azure infrastructure provisioning, Kubernetes container orchestration, and GCP Apigee hybrid runtimes under a single git-driven workflow.
Defined Enterprise SLA Targets: Built a production-ready template capable of meeting strict business SLAs (RPO < 15 mins, RTO < 1 hour, 99.99% availability target).
What we learned
Deep insights into managing hybrid multi-cloud control planes (GCP Apigee) operating on heterogeneous infrastructure (Azure AKS).
Best practices for designing progressive CI/CD pipelines that balance developer agility in sandbox environments with strict audit and compliance controls for Production.
What's next for Mahesh Project
Multi-Region Active-Active DR: Expanding the Terraform architecture to support multi-region Active-Active failover across distinct Azure geographies.
Agentic AI Gateway Integration: Exposing Apigee proxy flows as Model Context Protocol (MCP) endpoints to allow autonomous AI agents to securely interact with backend enterprise APIs under strict rate-limiting and OAuth controls.
Log in or sign up for Devpost to join the conversation.