How the cloud native Edge is transforming business at The Home Depot

If I was to walk into a store, what would I see behind the scenes? What does the infrastructure look like in terms of analytics, inventory, point of sale? Talk to us about that a bit.

For us, Edge spans a wide range of needs and applications — from stores, to distribution centers, to remote offices. Our strategy is to deliver a consistent way for applications to get to the Edge, and really provide innovation and the ability to deliver our interconnected retail experience. So, if you look at a Home Depot store, anything from online pickup in store, to inventory, to POS is all supported with Edge compute.

The Home Depot is doing great things at the Edge. Is it my understanding that you've also standardized on Kubernetes and cloud-native approach for all things Home Depot?

Absolutely. The runtime is centered around Kubernetes with some exterior components that allow us to present easy integration for our application teams, developers, our associates that support stores, and to improve customer experience.

Excellent. From our side, we see a lot of Edge opportunities globally. And so, we’ve also seen consistent challenges when it comes to the scale of these deployments. As you know, if you've got Kubernetes without the right tooling, it can be very challenging to manage. Deploying Kubernetes for a footprint of over 2,200 stores is very impressive. What key challenges and pain points have you experienced with this particular deployment?

If I look at the top three challenges, scale has to be No. 1. Our Edge environment is large and presents several unique challenges. However, we see having a common platform with Kubernetes at the Edge, centered around a single binary experience with K3s, enables us to deploy, update and secure the underlying container platform quicker than traditional full-stack Kubernetes.

Secondarily, resiliency is top of mind for us. With Edge compute, consistent communication is never guaranteed. Providing a resilient Edge-based runtime allows us to increase resiliency around those areas. As Home Depot, we operate in some areas where natural disasters happen, and our core value drives us to give back to our community we serve. Having to manage a platform in K3s allows us to rest easy that we'll be available even in disconnected scenarios, ensuring workloads are running smoothly.

You've almost preempted my next question. Over the years, in the cloud space, we've learned that simplicity scales. So, we've embraced that level of thinking at SUSE, and you can see that has materialized in the form of K3s, which is our lightweight Kubernetes engine. Tell us about anything additional for K3s, and possibly describe the technology stack that you're using.

Typically, across the Edge, we run a small high availability (HA) cluster setup. As you mentioned, it's roughly 2,200 clusters we operate in production at scale. K3s provides us a few different experiences. One is the HA availability or resiliency that we need for our business.

There is also a stand-alone mode at the Edge. The stand-alone mode allows us to build immutable, on-demand quality assurance (QA) clusters for our developers. It helps us to serve multiple use cases, from QA and testing, all the way through production workloads and resiliency.

Excellent. Something else we've noticed is that when we entertain these opportunities for Edge deployments, customers have varying levels of maturity. The Home Depot has a very sophisticated approach, and there's standardization across all your locations. We are keen to meet customers where they are on their journey to scale. Looking forward, what’s your vision for the next 10, five years, or even the next six months?

The technology landscape moves fast, so I don't know if I can predict the next month versus the next five years. What I do see is additional features and Kubernetes being released that makes ease of adoption that much easier. We're abstracting more and more infrastructure components, which build speed to market.

And then we're also seeing Internet of Things (IoT), artificial intelligence (AI) and machine learning (ML) starting to look at Edge use cases. So that's going to unlock a lot of features as we move forward. We're happy to see what that will provide us in the next several years.

Ultimately, day to day, we're focused on improving the customer experience and experience of our store associates.

That's a great point. One thing that I want to key in on is that, by us having this kind of standardized cloud-native Kubernetes infrastructure, it allows us to get that speed to market — that agility going forward — and to bring on these new applications in a competitive situation. I think The Home Depot sets an example in the space. I'm so proud of the team and the technology that we are bringing to the table to help you win. Is there anything that we can do to help you going forward?

I would say keep focusing on the gnarly problems at the Edge. SUSE Rancher and K3s solved a ton of those for us. Also keep flexibility. As we mentioned, we use K3s in a few different ways to solve needs for our business, and having that flexibility is key for us solving our business problems.

Thank you for that. We'll take that advice. I want to thank you and the team behind The Home Depot Edge deployment. You are rockstars, we love working with you, and we wish you the best of success in the world.