Deciding where to locate compute power is essential to ensure vital IT systems run efficiently with maximum uptime.
It’s a challenge that many organisations face in today’s always-on economy. As the retail sector becomes increasingly dependent on a mix of resource-hungry applications, IT infrastructure must function effectively and reliably in real-time and often at the edge.
Adding further complexity is the diversity of technology underpinning retail systems. Some were built on legacy infrastructures, while others are modern, lean platforms powered by AI. This can include anything from point-of-sale (POS) and ecommerce platforms to video monitoring, marketing, customer support, financial and analytics platforms. Often, they need to knit together to deliver a seamless customer journey across multiple touchpoints — physical stores, online platforms and mobile channels.
What makes this yet more challenging for retailers is the fragmented nature of how they operate. And, in today’s fiercely competitive landscape, many processes need to perform at speed with 24/7 uptime, while also protecting customer data from cyberattacks. It’s a multi-faceted issue that has been notoriously difficult to address.
However, the advent of virtual machines (VMs) marked a turning point. Running VMs on-premise have been helping to address the old stumbling blocks of latency and security, encouraging retailers to transition from traditional IT approaches to solutions that locate compute resources at the point of need. These alternatives have been able to offer some improvements in performance and resilience in edge environments.
A step onwards from virtualisation
As a result, virtualisation has become commonplace in retail IT infrastructures, whether using VMs or containers. It’s not hard to see why, as both containers and VMs enable applications to function independently and securely on the same hardware. For example, a single physical machine could run POS software, inventory and security applications. Each can coexist and receive separate software updates without causing any issues to the other.
Until recently, the preference has been for VMs, but the need for more responsive edge computing is driving the uptake of multimodal containers – deploying infrastructure that can run both VMs and containers. It’s a more powerful and adaptable way of using containers across different environments, including inside VMs, at the edge, or in the cloud. This hybrid approach gives businesses more options for modernisation without breaking what already works or immediately replacing legacy systems.
It means multiple containers can run on the same hardware and different operating systems, maximising performance and ensuring business continuity if cloud connectivity is unstable. Importantly, containers are not restricted to running natively on an organisation’s core operating system alongside a separate hypervisor running the VMs. Instead, applications can run inside
containers that are running within a VM, helping to reduce infrastructure overheads. Crucially, processing workloads where the data is created also dramatically reduces latency and improves performance in remote locations, which in turn increases customer satisfaction and maintains competitive advantage.
This hybrid way of working is also popular with IT teams as many are already adept at managing VMs, whereas containers require different in-depth expertise which may not be readily available in-house. In the not-too-distant future container orchestration and monitoring will no doubt become more straightforward, but in the meantime, a hybrid model offers the best of both worlds by providing flexibility and resilience – both critical for retail environments.
High impact applications to prioritise
By taking this approach, retailers can quickly deploy apps within stores. A host of applications quickly spring to mind to begin the transition, including:
● Point-of-Sale (POS). Containers can manage transaction processing across multiple store locations, optimising speed and uptime as well as facilitating security and software updates.
● Inventory Management. Edge-based containerised applications can provide real-time monitoring of stock levels and accurate inventory forecasting. This helps to improve demand planning and mitigate supply chain complications, therefore reducing stockouts and increasing sales.
● Customer-Centric Marketing. A containerised approach gives customers a consistent experience across all channels by integrating back-end processes for IoT, mobile and stores. This supports omnichannel marketing, ensuring pricing, promotions and fulfilment work seamlessly for the customer regardless of the way they choose to engage.
● Video Monitoring. Enabling real-time AI-driven video processing and analytics requires significant processing power which containers can accommodate at the edge. This helps stores improve loss prevention and strengthen overall store security. Useful insights regarding customer behaviour can also be gathered to enhance the local, in-store customer experience.
● Storage Efficiency. Processing data onsite instead of transferring data to the cloud saves energy costs and reduces expensive bandwidth requirements too.
Fivefold growth in containers predicted
For retailers looking to implement multimodal containers as part of an edge computing strategy, it makes sense to start small by deploying containerised applications in high-impact areas like POS and inventory management and then scale according to priorities.
Implementing the containers inside VMs and evolving towards native deployment will provide the interoperability needed to gradually decommission legacy systems without disrupting day-to-day operations. At the same time, this strategy allows innovative new services and functionality to be integrated rapidly within a controlled and secure environment.
With the container market predicted to grow more than fivefold by 2030, retailers should be seizing this opportunity to modernise with a more adaptable, economic approach to managing distributed workloads. It’s a way forward that will enable retail businesses to increase agility and manage costs effectively, while delivering a better, faster, more reliable customer experience.
About the Author
Bruce Kornfeld is Chief Product Officer at StorMagic. Serving organizations of all types and sizes, we help store, protect, and manage their data at and from the edge. Our products ensure data is always protected and available, no matter the type or location, to provide value anytime, anywhere. Our products operate from the harshest edge environments to pristine datacenters, from small businesses to Fortune 500 companies. They are simple, flexible, reliable, and cost-effective, without sacrificing enterprise-class features, for organizations with one to thousands of sites. Across many industries, including retail stores, factories, hospitals, universities, and even wind farms and oil rigs, we are helping organizations with their digital transformation.
