Book a demo

Prism · Standalone appliance

Build Kubernetes objects at scale, on a single node.

Prism is a standalone yet integrated appliance that lets customers and partners build Kubernetes and Isovalent objects at scale, all on a single Kubernetes node. The node takes on multiple personas, carries the data-source CRDs, and appears in Fabrix as a real data source. No cluster fleet, no lab build-out.

A whole cluster estate on one node

One appliance, five switchable personas.

  • cilium-oss
  • isovalent-k8s
  • isovalent-openshift
  • ovn-kubernetes
  • ovn-openshift

Scale objects with a slider. Bulk create with Batches.

What Prism gives you

Model the estate, not just the object count.

  • OVN-Kubernetes personasSwitch the node between ovn-kubernetes and ovn-openshift to model OpenShift's default CNI, including OVN egress firewalls.
  • Application tier modellingTie back into Fabrix to pull applications defined on the canvas and scale those tiers up on the appliance. Model a real app shape, not just object counts.
  • Batches or a sliderBulk create objects to your own naming conventions and label sets, or skip the detail and simply scale with the slider. Labels become real Cilium identities.
  • Up or down in secondsDial nodes, pods per node, namespaces and KubeVirt VMs and the estate resizes in seconds. Scale up for a test, scale straight back down when it's done.
  • Saved configurationsImport and export saved configurations to stand up different scenarios on demand: pre-migration testing and what-if modelling, repeatable across teams.
  • A real Fabrix data sourceGenuine CRDs (cilium.io, isovalent.com, config.openshift.io, kubevirt.io) with policy at volume: CNP, CCNP, INP, ICNP, encryption, CIDR groups and pod networks.

On the roadmapILB: integrated load balancing for Isovalent modelling. Extends Prism from endpoint and policy scale into load-balanced service modelling.

In the Fabrix console

For you
System
PrismA whole cluster estate on one node.
Target-scale clusters on one nodeScale with a sliderRehearse a migration at scale