Skip to content

Getting Started

The MetalSoft CloudPlex Platform is a powerful Software Defined Datacenter (SDDC) platform, with various modules that can handle many different use cases. Not all modules are used at the same time by all customers. The most common use cases for the MetalSoft platform are the following:

  • Full Stack deployment (servers, switches, storage, clusters etc.) In this setup all of MetalSoft functions are used in conjunction with one another.
  • Fabric Manager (network only management) In this setup only the network part is used.
  • Bare-Metal-as-a-Service (physical compute and storage but with externally managed network) In this setup the clients consume the Bare-Metal-as-a-Service via CI/CD or the client user interface but network provisioning is done outside MetalSoft either coordinated externally or via MetalSoft workflows.
  • GPU-as-a-Service (A variation on the Full Stack Deployment or the BMaaS) In this setup the clients consume the Bare-Metal-as-a-Service (BMaaS) and/or VM-as-a-Service (IaaS) via CI/CD but network provisioning is done outside MetalSoft either coordinated externally or via MetalSoft workflows.
  • VM as a Service - IaaS (VM cluster deployment and VM management) In this setup the consumption is primarily via VMs rather than Bare Metal nodes. MetalSoft handles cluster provisioning and lifecycle, potentially across multiple hypervisors. The networking part is done via network provisioning (overlay networks are implemented at the switch layer rather than in software).

The MetalSoft CloudPlex Platform is modular and some but not necessarily all modules are used in various use cases depending on requirements and existing automation.

Use caseFabric ManagerServer & Storage ManagerVM ManagerCluster Manager
Full Stack deploymentyesyesyesyes
Fabric Manageryesnonono
Bare Metal-as-a-Serviceyesnonono
GPU-as-a-Serviceyesyesyesno
VM-as-a-Service (IaaS)yesyesyesno

Depending on what you are using it for you may find that some materials are more important than others. Nevertheless we recommend starting with the the overall architecture and then moving on to learn about the various aspects related to the respective use case.