SpectroCloud — Kubernetes & VM Management
Hands-on SpectroCloud Palette — declarative Kubernetes from edge bare metal to the core datacenter, running containers and VMs side by side.
Not available to book
This demo is under maintenance and cannot be booked. It will reopen once its owner marks it ready.
What a customer gets out of it
- Show Kubernetes managed declaratively — the full stack from OS to add-ons defined as a versioned profile, not assembled by hand
- Run VMs and containers side by side on one platform, so migration and refactoring can happen at the customer's pace
- Demonstrate the same operating model reaching from edge bare metal to the core datacenter
- Explore AI workloads on Kubernetes — GPU scheduling and inference on the same platform as everything else
- Support like-for-like comparison against VMware VCF, Nutanix, OpenShift and HPE VME on identical lab hardware
Why it matters
Most organisations end up running Kubernetes in several places at once — a datacenter cluster, something at the edge, something in a public cloud — and each is built and upgraded differently. That divergence is where the operational cost actually lives, and it is what stops Kubernetes being viable for mission-critical workloads. Palette's answer is to make the whole stack declarative: the OS, the Kubernetes version, CNI, CSI and every add-on are defined in a versioned Cluster Profile and applied identically wherever the cluster runs. Because it also runs VMs alongside containers, a customer does not have to finish a refactor before they can consolidate — they can host what they have today and modernise incrementally. For organisations weighing a hypervisor renewal against a container platform, that combination is often the more cost-effective route to the modern datacenter, and it is testable here against the alternatives on identical hardware.
Lab overview
No walkthrough is recorded for this demo yet, so this is a 21 sec tour of the lab and how to use it.
Trouble playing? Download the file.
Overview
This environment builds a SpectroCloud Palette managed Kubernetes cluster in one of the lab's demo landing zones, so customers, SMEs and engineers can get hands-on with declarative Kubernetes — and with running virtual machines and containers on the same platform — without touching a production estate.
Bookable as a guided demo, a hands-on POC, or a classroom for a group.
Why Palette
Kubernetes rarely stays in one place. A typical estate ends up with a cluster in the datacenter, something running at the edge, and something in a public cloud — each built by different hands, on different versions, upgraded on different schedules. That divergence is where the operating cost lives, and it is the reason many organisations hesitate to put mission-critical workloads on Kubernetes at all.
Palette's answer is to treat the entire stack as declarative. A Cluster Profile defines the operating system, the Kubernetes version, CNI, CSI and every add-on as versioned layers — and the same profile is applied whether the target is a bare-metal box at a retail site or a cluster in the core datacenter. Clusters stop being artefacts someone assembled and become something reproducible.
What you'll explore
- Cluster Profiles — the declarative, layered stack model. Change a layer, and the platform reconciles every cluster built from that profile. This is the differentiator, not a feature.
- Palette VMO (Virtual Machine Orchestrator) — run VMs directly on Kubernetes alongside containers. The path for customers who need to host existing workloads now and refactor later, rather than finishing a migration before they can consolidate.
- Edge — the same operating model out to bare metal at the edge, with low-touch provisioning and an immutable edge OS for sites that have no one on hand to fix them.
- AI workloads — GPU scheduling and inference running on the same platform as the rest of the estate, rather than a separate stack nobody else can operate.
- Day-2 at scale — upgrades, patching and drift detection driven from the profile, across every cluster at once.
Edge to core, one operating model
The edge case is genuinely different: constrained hardware, unreliable networks, no local staff, and often far more sites than the datacenter has clusters. Palette treats those sites with the same declarative model as the core — which means one way of building, upgrading and auditing a cluster, whether there are three of them or three hundred.
For a business running mission-critical workloads across that spread, the value is not any single feature. It is that the edge estate and the datacenter estate stop being two different operational problems.
What you'll learn
- Building a cluster from a declarative profile instead of assembling one by hand
- Running VMs and containers together, and what an incremental refactor actually looks like
- How the same profile behaves against edge bare metal and against datacenter hardware
- Scheduling AI and GPU workloads on a general-purpose platform
- How Palette compares against VMware VCF, Nutanix, OpenShift and HPE VM Essentials — all of which run on identical lab hardware, so the comparison is like-for-like