I’m evaluating a hybrid API management architecture using Apigee Hybrid and would like to clarify a point regarding platform support vs Kubernetes compatibility.
From the official documentation, I understand that Apigee Hybrid is certified for specific platforms (GKE, EKS, AKS, etc.), but fundamentally relies on Kubernetes (with supported versions and Istio).
In this context, I’m considering running the runtime plane on Oracle Kubernetes Engine (OKE), which provides a CNCF-conformant “vanilla” Kubernetes environment.
Given that:
OKE is standard Kubernetes (no major vendor-specific abstractions)
It supports compatible Kubernetes versions and typical ingress/Istio setups
Apigee Hybrid is designed to run on Kubernetes clusters
My questions are:
Has anyone successfully deployed Apigee Hybrid on OKE (OCI)?
Are there any known limitations or blockers compared to officially supported platforms?
From a support perspective, would this be considered a “best-effort” scenario?
Any recommendations around networking (especially control plane connectivity to GCP) or ingress setup in this type of environment?
Appreciate any real-world experiences or guidance.
Welcome to the Apigee community @acer we’ll continue to monitor this thread and invite other members to share their insights. In the meantime, please review the resources below, which I hope will be helpful to your question:
I’d also love to invite you to our Community Tech Talk this week on April 16 at 5 PM CEST. It’s a fantastic opportunity to connect with Googler and community experts live:
Powering Retail Agents with ADK, UCP & Apigee X |Register here!
I have a reasonable perspective of active Apigee hybrid deployments in the wild and am not aware of any deployments on OKE.
No, not really. As far as what the Apigee hybrid runtime is and requires, it’s all pretty standard stuff. The need to install CRDs, an operator, and the requirement for PersistentVolume resources is about as advanced the requirements get.
Yes, you could call it “best-effort”, however, Google Support reserves the right to fallback to a zero or minimum level of best-effort support if the product is not being operated within the official parameters clearly instructed and stated in documentation.