Edge computing sprawl is an emerging problem. One company has a possible solution.
Edge computing is at the peak of the hype cycle right now. In simple terms, it refers to computing devices and applications at the network edge. ZK Research defines this edge as anything that is not a centralized compute node; it includes campus, branch, Internet of Things (IoT), 5G and numerous emerging “edges.”
Companies today already struggle to manage data and resources in the cloud and are now putting even more data in more distributed locations. Without the right management tools, business will quickly lose control over their infrastructure as edge grows and gives rise to “edge computing sprawl.” This is similar to the challenge virtualization had in its early days, prompting VMware to build vCenter.
Scale Computing: an Early Edge Computing Provider
Scale Computing is well-versed in the challenges of edge computing. The company got its start as a hyperconverged infrastructure (HCI) provider, which encompasses elements of a traditional data center, including storage, compute, networking, and management.
One could look at HCI as an early version of the edge computing model. The vendor has taken its experience in this area and used it to build a management platform to help its customers scale edge deployments without getting bogged down in complexity.
In my latest ZKast interview, I talked to Craig Theriac, VP of Product Management at Scale Computing, about how the company is helping businesses manage their edge infrastructure with its new cloud-hosted Fleet Manager tool. Highlights of the ZKast interview, done in conjunction with eWEEK eSPEAKS, are below.
- The idea of running computing close to users has existed for a long time, but edge computing as a concept has only taken off recently. With the rise of higher performance graphics processing units (GPUs) and data processing units (DPUs), as well as flash storage becoming less expensive, edge computing has been growing at an astonishing rate.
- One of the biggest drivers of edge computing is companies needing to consistently rely on their Internet connection. Regulation is another big driver. Depending on the industry and laws in specific regions, companies need to keep personally identifiable information (PII) local. The last driver is cost. While cost is coming down, it’s still a concern for many companies.
- Scale Computing infrastructure has historically been used for mission critical workloads. It provides a virtualization layer alongside a storage layer and uses a hypervisor that is a KVM-based. The storage layer is designed to be consumed by a KVM hypervisor. This enables it to run on small form factor compute nodes.
- Scale Computing recently launched Fleet Manager, a comprehensive tool that manages and monitors the health of distributed IT infrastructure. Fleet Manager shows real-time conditions for a fleet of clusters, including storage and compute resources. So, IT teams can centrally oversee deployments—whether it’s one or 50,000 clusters—and quickly identify problem areas.
- Fleet Manager is deployed alongside Scale Computing’s edge platform HyperCore. The introduction of Fleet Manager on HyperCore is what makes a solid edge computing solution. HyperCore has intelligence built into it, which allows it to monitor thousands of conditions. For companies with many individual sites, Fleet Manager adds another layer for managing a fleet of HyperCore base clusters.
- Companies with distributed environments often don’t have IT resources onsite and experience hardware failure. If problems occur, Fleet Manager prioritizes what needs to be fixed and sends alerts to administrators to notify them about that site. In a retail scenario, for instance, a system can stay online even if there is a node failure.
- Retail is a big industry for edge computing. If a point of sale (POS) system goes down or a customer loyalty program isn’t working, shoppers abandon their carts and hurt the retailer’s bottom line. The retail industry requires some level of autonomy to run infrastructure on-prem, especially in rural areas.
- Ahold Delhaize is a multibillion-dollar Belgium-based grocery retailer, with thousands of stores worldwide. It initially deployed HyperCore, but later needed an overlay as the retailers continued to expand. Scale Computing tailored Fleet Manager to the retailer’s specific use case. This has allowed Ahold Delhaize to effectively manage many clusters across hundreds of stores, while saving time and money.
- Scale Computing also recently launched zero touch provisioning with local USB support for collecting configuration information from nodes. Next on Scale Computing’s roadmap is to have the configuration information directly available within Fleet Manager, which would replace manual processes.