DEVOPSdigest

Spectro Cloud announced its Palette Edge platform, extending the Palette Real Metal architecture released last year. The offering sets a new industry standard for comprehensive management of Kubernetes on edge locations at the scale businesses require.

Spectro Cloud’s core Palette platform is a modern approach to Kubernetes management, enabling organizations to easily manage the complete full stack of any combination of new or existing K8s cluster. Palette goes beyond the mere K8s infrastructure. It includes the operating system and application services across virtualized or bare metal data centers, public clouds and now edge locations.

Palette differentiates Spectro Cloud from its competitors by addressing the challenges of today’s collection of disaggregated open source and commercial technologies and offerings for the edge through providing a next-generation platform that delivers centralized and automated Kubernetes lifecycle management. The platform’s architecture is based on decentralized orchestration and policy enforcement to enable virtually infinite scale without affecting performance.

Launched last year, Palette’s Real Metal architecture provides an integrated set of features, optimized to address the unique challenges of managing bare metal environments in conventional and “near edge” data centers. The architecture leverages open source technologies, such as CNCF’s Cluster API, Canonical “Metal-as-a-Service” (MAAS) and Spectro Cloud’s open-sourced Cluster API MAAS provider.

Palette Edge is a result of extending the core Palette platform and its Real Metal architecture to meet the specific needs of edge solution delivery and management.

It directly addresses the lack of an effective commercial solution, and offers cost-efficient, centralized management at scale for unsupervised edge locations, based on heterogeneous K8s distributions that run on bare metal single-server configuration. The solution significantly reduces the IT skills required for organizations to deliver modern applications at the edge.

Palette Edge addresses these challenges by enabling IT and DevOps teams to:

- Add edge devices easily and quickly with a “plug-and-play” approach;

- Scale to virtually any number of K8s edge locations, without limits or performance deprivation, due to local at-cluster enforcement of policies;

- Perform rolling upgrades even with single-server edge configurations, with no impact to the availability, due to a unique failsafe, zero-downtime capability;

- Run both containerized and virtualized workloads on x86 and ARM-based edge devices; and,

- Deploy pre-validated and curated K8s stacks comprising any combination of distribution, operating system and add-on application services, while managing them consistently across day 0, day 1 and day 2 operations via a single control point and cloud-like experience.

“Enabling developers and operations teams to centrally, and therefore consistently, deploy and operate Kubernetes application stacks at edge locations with Spectro Cloud’s heterogeneous server, storage and network equipment and configurations is a big deal,” said Torsten Volk, managing research director at Enterprise Management Associates. “Palette Edge brings declarative full stack application management to the edge, but leaves the heavy lifting to the Kubernetes edge clusters themselves. This enables organizations to push out self-managed Kubernetes apps, with infrastructure teams managing centralized cluster profiles that tell each Kubernetes cluster what to do. This is very exciting, as it constitutes the foundation for the mass adoption of edge Kubernetes that could open up many new use cases that could significantly move the needle for many organizations.”

Palette Edge is purpose-built to support today’s edge use cases such as Internet of Things device management and orchestration, data ingestion, streaming, analytics and AI inference to support a variety of industries including telecommunications, hospitality, retail, food and beverage, energy, and oil and gas and health care.

“Bare metal K8s in the data center and at the edge are use cases that share the same main concept, but have important differences,” said Spectro Cloud co-founder and CEO Tenry Fu. “Edge for our customers often means single-server, unsupervised locations with limited connectivity. We designed an architecture that can now treat edge locations as almost autonomous, with local 'at cluster' policy enforcement, rolling updates and support for any workload. This significantly reduces the operational risk and cost, while also enabling scaling to thousands of clusters without diminishing performance. No other solution does that.”