In modern banking, money might be the product, but data is the power behind the machine. And when infrastructure is juggling millions of transactions and trillions of dollars in payments, Kubernetes data management becomes critical and resilience isn’t just a goal — it’s the mandate.

At HSBC Group, that mandate takes the form of a Kubernetes platform operating at an extraordinary scale. With 13,000 nodes, 200-plus clusters and more than 600 production services, the system is under pressure to perform with uncompromising reliability and zero tolerance for disruption. The bank’s infrastructure demands have become so extensive that platform decisions can make or break national economies, according to Steve Lewis (pictured, left), global head of engineering for container platforms at HSBC Group.

Pure Storage’s Venkat Ramakrishnan and HSBC’s Steve Lewis talk with theCUBE about the complexity of operating Kubernetes at scale.

“We are looking for very high reliability,” he  said. “We cannot take any risks to those platforms. The throughputs are huge. We’re talking about millions of transactions and trillions of payments, reliability and integrity of the data, and our ability to survive any kind of incident and know that the data is intact is absolutely critical.”

Lewis and Venkat Ramakrishnan (right), vice president and general manager of Portworx by Pure Storage Inc., spoke with theCUBE’s Paul Nashawaty at KubeCon + CloudNativeCon Europe, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the complexity of operating Kubernetes at scale, HSBC’s journey with Portworx, and the future of self-service platforms in data-driven enterprises. (* Disclosure below.)

HSBC scale smarter with Kubernetes data management

Performance is only the beginning when it comes to managing infrastructure at HSBC. The real challenge lies in navigating a tangled web of regulations, scale demands and relentless upgrade cycles. With Kubernetes clusters sprawling into the hundreds and workloads spread across continents, the bank needs a platform that can do more than simply keep up. It needs one that can evolve with it, according to Lewis.

“The constant flow of upgrades — new versions of Kubernetes are coming out every four months — and we’ve got to keep pace,” he said. “We’ve got to get that upgraded. Normally, a state the size that we have — 200 plus clusters — that’s like painting the Forth Road Bridge; it never ends. We are constantly upgrading [and] maintaining. I think that’s where … one of those value-added services we’ve been able to leverage from Portworx is the Portworx backup feature.”

Those backup capabilities have become essential for recovery and the starting point for something bigger. Portworx gives HSBC the flexibility to leapfrog versions and meet regulatory demands, offering a broader Kubernetes data management feature set. Features such as disaster recovery, replication and vendor-agnostic deployment have made it easier to manage Kubernetes at scale, according to Lewis.

“You need to rewrite many [capabilities],” he said. “That was the basic reason we got into Portworx in the first place. But that’s just if you like [the] price of entry: It’s the basic requirement. Once you’ve got there, there’s more value in the platform itself. You’ve got the backup capabilities, the DR capabilities, the replication and all that kind of stuff. It’s a platform on which we can build.”

For Portworx, enabling that kind of scale means embedding deeply into both the control plane and the enterprise culture. Supporting thousands of nodes and dozens of teams requires more than tooling. Scalable Kubernetes data management also demands adaptability, according to Ramakrishnan.

“We integrate with the customers’ processes,” he said. “That’s the key thing. A lot of folks have built bespoke or products for specific use cases, but the customers have to chain their workflows to integrate them. What we do in Portworx is that we work with the customers’ workflow and integrate our product into theirs, so it’s easy for them to adopt and operate at scale.”

That integration has been key to improving Kubernetes data management across HSBC’s global footprint, according to Ramakrishnan. Along the way, the bank has gained a more consistent and developer-friendly experience across its container ecosystem.

“We are built as a fully integrated piece of the Kubernetes control plane. We run in the worker nodes; we give a Kubernetes experience to all of the operators and all of the folks who operate the clusters,” Ramakrishnan said. “The developers also get the complete Kubernetes experience of self-service. It’s easy to deploy and scale as well.”

Photo: SiliconANGLE