Edge Clusters provide you with the tools needed to dynamically grow your cloud infrastructure with physical or virtual resources running on remote cloud providers. Edge Clusters support two main use cases:

  • Edge Cloud Computing, to transition from centralized clouds to distributed edge-like cloud environments. You will be able to grow your on-premises cloud with resources at edge data center locations to meet the latency, bandwidth or data regulation needs of your workload.
  • Hybrid Cloud Computing, to address peaks of demand and need for extra computing power by dynamically growing your underlying physical infrastructure.

How Should I Read This Chapter

In this chapter you can find a guide on how to automatically allocate and provision Edge Clusters on bare-metal and virtual instances on cloud providers:

In this chapter you’ll also learn how to operate your clusters and manage providers.

Hypervisor Compatibility

Edge Clusters can be virtual or metal depending of the instance type used to build the cluster. Note that not all providers offer both instance types.

Edge/Cloud Provider Edge Cluster Hypervisor
Equinix metal KVM, Firecracker and LXC
AWS metal KVM, Firecracker and LXC
AWS virtual LXC
Google virtual LXC
DigitalOcean virtual LXC
Vultr metal (in progress) KVM, Firecracker and LXC
Vultr virtual LXC
On-prem metal KVM, Firecracker and LXC

The Edge Cluster type determines the hypervisor and workload that can be run in the cluster. The following table summarizes the Edge Cluster you need to run specific workloads:

Use Case Edge Cluster Hypervisor
I want to run application containers… virtual LXC
metal LXC, Firecracker
I want to run virtual servers… metal KVM, LXC
I want to run a Kubernetes cluster… metal KVM (k8s based)
Firecracker (k3s based)

In the above table, application containers are those imported from DockerHub, LinuxContainers or TunrkeyLinux, as well as images created from DockerFiles. On the other hand Virtual servers use full system disk images.