We invited Chris Hendrich, the Associate CTO at SADA and a Google Cloud Certified Fellow, to give a Partner Talk on Google Distributed Cloud (GDC) resiliency in our booth at Google Cloud Next 2025.
Hendrich is an expert on resiliency at SADA, a global leader in providing business and technology consulting. SADA is a Google Cloud Premier Partner and has been named Google Cloud Global Partner of the Year eight years in a row.
Given all the technology and architecture involved in making Google Distributed Cloud (GDC) deployments resilient, his talk focused on the importance of the storage layer.
According to Hendrich, storage is the one critical layer many edge deployment teams still overlook in their resiliency plan. "Customers often think the network alone can solve everything," he said. "But you can't forget about storage. You need that storage layer to build a true resilient edge platform that you can roll out at scale."
Hendrich described how enterprises have embraced GDC for its flexibility and consistency. GDC enables the same orchestration and application management experience across cloud, on-premises, and remote environments.
However, as companies scale from pilot environments to real-world production with hundreds or thousands of edge sites, challenges surface that go beyond compute orchestration. One of the most urgent is resiliency.
Edge environments are inherently vulnerable to disruption. Power loss, accidental cable disconnections, or hardware failures can bring down a single-node GDC system entirely. This becomes particularly problematic in time-sensitive environments like manufacturing plants, where downtime is not an option.
“A lot of times assembly lines run 24/7,” Hendrich said. “If you have a single GDC node that needs to be updated and suddenly you have to reboot it. That takes five minutes to come back, and the assembly line can’t just stop for five minutes.”
To mitigate these risks, Hendrich recommended running a three-node GDC architecture to provide failover protection at the Kubernetes control plane level.
This architecture falls short of fully addressing application-level availability, especially for workloads that rely on persistent data.
Using its GDC design expertise, SADA can architect a highly reliable GDC network. But if a node running a VM goes down, that workload doesn’t simply restart on another node. All applications—but particularly legacy VM-based NFV applications—require persistent storage to maintain state, configurations, and operational continuity. Without a resilient and distributed storage layer, edge workloads remain fragile.
To deliver the persistent storage, SADA specifies Rakuten Cloud-Native Storage, which brings enterprise-grade storage capabilities to the edge. Rakuten Cloud-Native Storage is integrated with GDC and provides the missing foundation for data resiliency. Hendrich describes Rakuten Cloud-Native Storage as the "go-to solution" for customers building distributed edge architectures.
“What I was really looking for before we had Rakuten Cloud-Native Storage for GDC was this three-way replication,” he says. “You have three nodes. Ideally, you want your pod or your VM replicated across all the nodes. That’s all done really easily with minimal configuration with Rakuten Cloud-Native Storage.”
The storage software is designed for ease of deployment and scalability. Available through the Google Marketplace, it can be managed using the same Google Kubernetes Engine (GKE) or GDC console already in use.
This minimizes operational overhead, an essential requirement for organizations managing thousands of edge locations. Crucially, Google also supports Rakuten Cloud-Native Storage at the L1 and L2 levels, providing a single-vendor model for both compute and storage.
For edge use cases like AI inference, where data must be processed with low latency close to the source, Rakuten Cloud-Native Storage ensures that workloads stay running even if individual nodes go offline. It also supports traditional VM-based applications, enabling a hybrid transition to cloud-native architectures without breaking existing operations.
As customers increasingly explore modern cloud-native application platforms, they face complex migration scenarios. Not all applications can be containerized immediately, and many continue to rely on commercial off-the-shelf software that mandates VM support. Rakuten Cloud-Native Storage bridges this gap with persistent storage that works seamlessly with both containers and VMs.
With solutions like Rakuten Cloud-Native Storage and the flexibility of Google Distributed Cloud, enterprises now have the tools to build robust, scalable edge platforms that keep data safe, applications available, and operations running—no matter what happens on the ground.
Watch Hendrich’s complete presentation on YouTube: https://www.youtube.com/watch?v=1QBX0DXdsmw
.svg)
