Retailers, like everyone else at this point, have heard of the cloud; computing that is typically owned and operated by someone else, like Amazon Web Services, Azure and Google Cloud Platform. Cloud-native, on the other hand, is a newer set of technologies that enable scale, reliability and faster development velocity. If “the cloud” were the engine of a car, “cloud-native” would be nitrous oxide, racing tires and a pit crew rolled into one.
The most popular cloud-native technology is Kubernetes, a software-container orchestration platform that allows declarative configurations, replication and failover. Declarative configurations are key to building software delivery systems that are repeatable, auditable and scalable. Retailers especially will appreciate the elasticity of Kubernetes. When running an application on Kubernetes, an operator can set the number of desired replicas for that application and, as demand scales up, automatically add nodes to deal with additional traffic.
For example, on Black Friday a retail website going down or a payment processing gateway failing could cost millions of dollars in lost revenue. With Kubernetes we can easily scale the infrastructure for the burst of activity and then scale down afterwards.
Cloud-native is great for avoiding downtime and it’s incredibly impactful on the software development process. Most retailers are now recognizing that their ability to develop and ship software is key to their business success. With a cloud-native approach, software delivery cycles can drop dramatically. Not so long ago it was common for software development teams to only release new software once a month. Now, thanks to advancements made possible by cloud-native tech, teams are releasing multiple times per day. Delivering software faster means earning value on that software faster. Every retailer knows letting inventory sit on shelves costs money, but few realize that allowing their software changes to sit in staging is potentially costing them even more.
Even retailers that have embraced the cloud may be missing out on some of the advantages of cloud-native software. For example, cloud-native doesn’t mean the software only runs in the cloud. Many retailers are now using Kubernetes and other cloud-native technologies like Argo and Codefresh to provide services on location.
A Kubernetes cluster in a store location can provide fast services to the store. Where in the past running servers on-site was costly and maintenance difficult, cloud-native technologies allow for much easier and more reliable management. We have customers managing applications and services across hundreds or even thousands of locations, all while meeting high compliance and security requirements and greatly reducing management costs along the way.
The decision to go cloud-native may seem confusing and jargon-filled but it is nonetheless critical. When McDonald’s brothers developed the Speedee it probably seemed confusing and overly technical to their peers. We all know how that turned out.
Dan Garfield is a full-stack engineer, Kubernaut and Chief Open Source Officer at Codefresh, a DevOps automation platform. As a Google Developer Expert and a member of the Forbes Technology Council, he’s built open source tools for advanced deployment methodologies with Kubernetes, Helm and Istio. He’s helped companies small and large alike adopt Kubernetes and modern DevOps practices.