Four ways to run Kubernetes. One fleet, one governance.
Prime for production. Flex for everything around it. Edge for remote sites. Metal for your own servers.
Pick where your clusters run and how long they live
Same Kubernetes, same policies, same team running it.
Four ways to run PKE, one platform
Same Kubernetes, same policies, same team. What changes is where the clusters run and how long they live.
| PKE Prime | PKE Flex | PKE Edge | PKE Metal | |
|---|---|---|---|---|
| Where it runs | Pocteo Platform | Pocteo Platform | your sites | your data center |
| Nodes | dedicated to you | shared pool, sandboxed | your edge hardware | VMs on your servers |
| Isolation | dedicated kernel | gVisor, microVM for untrusted code | single tenant per site | single tenant |
| Connectivity | online | online | online, intermittent, air-gapped | egress-only link |
| Time to a cluster | sized with you | about a minute | boot from the image | after provisioning |
| Pricing | per node, monthly, in TND | capacity pools, in TND | per site | per server |
| Best for | production | CI, previews, tests, training | factories, vessels, stores | private, on-premises |
From pull request to production, on one platform
Govern them, build on them
Talk to an architect
We size production on Prime and previews on Flex with you, and bring your sites and servers into the fleet.