Skip to main content

Additional Capabilities

Palette offers a range of additional capabilities designed to enable its users to deploy, scale, and effectively manage Kubernetes workloads across a wide variety of environments and deployment options.

This section will introduce you to some of Palette's additional capabilities, which include:

  • Managing thousands of clusters in remote locations with Edge.
  • Self-hosting the Palette management plane in your own environment with Self-Hosted Palette.
  • Integrating virtual machine workloads into Kubernetes environments with Virtual Machine Orchestrator.

A drawing of Palette with humans interacting

Edge​

Palette Edge enables you to deploy Kubernetes workloads in remote locations characterized by limited or intermittent connectivity and limited compute infrastructure. This means you can deploy Kubernetes clusters at scale and ensure application performance, availability, security, and lifecycle management across a diverse range of edge locations. These locations include hospitals, retail stores, Telco environments, restaurants, manufacturing facilities, rural areas, and many more.

Palette Edge supports both VM and container-based workloads, multiple Kubernetes distributions, and Intel and ARM hardware architectures. It is built on top of the open-source project Kairos, which enables the creation and customization of immutable versions of operating systems. Additionally, Palette Edge is designed to scale to tens of thousands of locations while enforcing policies locally within each cluster.

Edge clusters are Kubernetes clusters set up on Edge hosts. These hosts can be bare metal or virtual machines located in isolated locations. Palette deploys and manages workload clusters at the Edge, and the services continue operating even when the connection to the management plane is lost. You can manage Edge clusters locally on-site through the local UI, or centrally through the Palette management plane. Palette Edge is able to meet your needs, regardless of the network topology your deployments face. Check out the Palette Edge page to learn more about Edge and its features.

Self-Hosted Palette​

By default, the Palette management plane is available as a multi-tenant SaaS deployment in a public cloud with multiple availability zones. Should you need it, Palette is also offered as a dedicated SaaS instance, as well as a fully self-hosted option that allows your teams to directly deploy and manage a private instance of the Palette management plane in your data center or public cloud provider.

Self-hosted Palette puts you in full control of the management plane, including its configuration and the timing of upgrades. A self-hosted instance may be necessary to meet compliance requirements or your organization's security policies. You may also need to deploy an instance of Palette within an airgapped facility to manage clusters where access to any outside service is not possible. Explore more on the Self-Hosted Palette page.

Virtual Machine Orchestrator​

Palette Virtual Machine Orchestrator (VMO) allows you to deploy, manage, and scale traditional VM workloads within a modern Kubernetes environment, side by side with your containerized applications. It lets you apply to VMs the same lifecycle management capabilities as Palette applies to containers, including backups.

VMO uses the CNCF project KubeVirt to manage VMs as Kubernetes pods, ensuring complete mapping between the VM and Kubernetes concepts. This solution also has near complete feature parity with VMware vSphere, including capabilities such as live migration.

Palette VMO can be used on edge hosts, giving the ability to deploy VM workloads at the edge without the overhead of a hypervisor layer. This is achieved by leveraging Canonical MAAS. Additionally, VMO can also be used in self-hosted, airgapped, and in our SaaS environments. Learn more on the Virtual Machine Orchestrator page.