01
Inventory
Metal, GPU pools, and sites that will host clusters.
Infrastructure · Fusionative Core
Control the infrastructure beneath your applications.
Kubernetes, VMs, KubeVirt/KVM, storage, networking, and bare metal under one operational model with Fusionative Core.
On this page
Fusionative Core infrastructure: Kubernetes fleets, VMs, storage, networking, and bare metal under one operational model.
Fusionative Core
Infrastructure under one operating model.
Kubernetes, VMs, storage, networking, and bare metal share tenancy and day-2 workflows, so AI cloud productization has a real substrate.
01
Fragmented estates often look like
- Separate consoles for K8s, virt, and storage
- Cluster create as a ticketed side process
- Network and SDS outside the AI cloud story
- No shared tenancy between infra and services
02
FusioNative Core brings
- Fleet provisioning and upgrades from one plane
- VM and KubeVirt/KVM beside containers
- Ceph-first storage and policy networking
- Bare metal in the same operational model
Must-have · K8s cluster provisioning
Lifecycle Kubernetes fleets from the control plane.
Stand up and operate clusters, including air-gap-friendly patterns, multi-cluster topology, HA and MAAS-backed provisioning paths, and day-2 upgrades under one plane of record. Built for sovereign and provider estates that cannot treat cluster creation as a side quest.
01
Provision and upgrade
Stand up and lifecycle fleets from Core, not a side console.
02
Multi-cluster / HA
Topology workflows for HA estates and provider-scale fleets.
03
Tenant isolation
Projects and RBAC on the same plane as the clusters.
04
Air-gap friendly
Patterns for restricted and sovereign install paths.
Cluster lifecycle
From capacity to day-2 fleet ops.
A simple sequence platform teams use when Kubernetes is the substrate for AI and multi-tenant clouds.
02
Provision
Stand up fleets with HA and air-gap-aware patterns.
03
Isolate
Tenants, projects, and quotas on the same plane.
04
Upgrade
Day-2 lifecycle without a separate cluster console.
05
Observe
Fleet health signals back into the operating model.
Infrastructure layers
Everything beneath the applications you run.
Six layers operators manage together when FusioNative productizes the estate into AI, NEO, sovereign, or air-gapped clouds.
01
Kubernetes
Provision, upgrade, and observe cluster fleets across environments.
02
Virtual Machines
Manage VM workloads alongside containers from one control plane.
03
KubeVirt / KVM
Run virtualized workloads on Kubernetes with unified policy.
04
Storage
Centralize lifecycle and visibility across storage backends.
05
Networking
Policy, DNS, load balancing, and service connectivity.
06
Bare Metal
Bring physical capacity into the same operational model.
Multi-cloud reach
One control plane. Multiple estate types.
Operate private, public, sovereign, edge, and bare-metal estates without fragmenting policy or visibility.
fusionative-core · control plane reach
Policy · Identity · Observability · FinOps
01
Private Cloud
Logical control-plane reach, not a vendor cloud map.
02
Public Cloud
Logical control-plane reach, not a vendor cloud map.
03
Sovereign Cloud
Logical control-plane reach, not a vendor cloud map.
04
Edge Cloud
Logical control-plane reach, not a vendor cloud map.
05
Bare Metal
Logical control-plane reach, not a vendor cloud map.
Storage · Fusionative Storage
Fusion Ceph first: block, file, object—plus Vault for backup.
Rook-operated Ceph is the primary SDS path (RBD, CephFS, RGW). Advanced site continuity uses RBD mirroring and RGW multisite. FusionVault covers backup, WORM, and air-gap—not Ceph peer replication.
01
Fusion Ceph
Wizard-driven Rook deploy, OSDs, pools, CephFS, RGW, and StorageClasses from the control plane.
02
Mirroring & multisite
RBD mirroring for block images between clusters; RGW multisite for object zones—named and scoped as Ceph provides them.
03
FusionVault
Backup and DR patterns (Velero/Kopia, WORM, air-gap). Complements Ceph replication; does not replace it.
Backend visibility
Ceph is first-class. Other backends appear as lifecycle and visibility integrations, not claimed replacements.
01Ceph
02Longhorn
03CSI
04S3-compatible storage
Networking · Fusionative Network
CAMP on Cilium—Calico and Kube-OVN when the estate needs them.
CAMP is the Fusionative Enterprise Cilium engine: eBPF connectivity, NetworkPolicy, Hubble, Tetragon signals, and Cluster Mesh. Calico and Kube-OVN stay supported alternative CNIs. Exposure Gateway and Fabric planning sit on the same Network product.
01
CAMP · Cilium
Enterprise networking engine: policy, Hubble, Tetragon adjacency, Cluster Mesh.
02
Calico
Supported CNI for estates already standardized on Calico NetworkPolicy paths.
03
Kube-OVN
Supported CNI for OVN-oriented underlays—operable from the control plane.
04
Exposure & fabric
DNS, certificates, routes, GitOps export, plus CIDR/MTU/topology planning.
Put your estate on one infrastructure plane.
Walk Kubernetes provisioning, storage, and networking with the platform team, then map it to the AI cloud you want to operate.