Orchestration without the PhD.
The Karbon Scale brings full workload orchestration to Tophan. One-click cluster provisioning, automated scheduling, and integrated networking — all managed through Dragon’s Eye or the CLI.
| Feature | Description | Status |
|---|---|---|
| 1-Click Clusters | Provision a complete orchestration cluster from Dragon’s Eye. Networking, storage, and DNS configured automatically. | Planned |
| Pod Networking | Integrated with the Networking Scale. Automatic overlay networking, DNS, and service discovery. No separate CNI to install. | Planned |
| Persistent Storage | Native integration with the Storage Scale. Volumes provisioned, attached, and replicated automatically. | Planned |
| Ingress | Built-in L7 ingress with TLS termination, path-based routing, and automatic certificate management. | Planned |
| GitOps | Declarative workload definitions synced from Git. Push to deploy. Drift detection and automatic reconciliation. | Planned |
| Private Registry | Integrated image registry backed by the Storage Scale. Content-addressable, deduplicated, signed. | Planned |
| Monitoring | Workload metrics, logs, and health status integrated with the Monitoring Scale. No separate observability stack to deploy. | Planned |
| Auto-Scaling | Horizontal pod scaling based on CPU, memory, custom metrics, or schedule. | Planned |
| Rolling Updates | Zero-downtime deployments with configurable rollout strategies and automatic rollback. | Planned |
| Multi-Tenancy | Namespace isolation with resource quotas, network policies, and RBAC. | Planned |
Traditional orchestration platforms are complex because they’re designed to run on any infrastructure. They can’t assume anything about the network, the storage, or the security model, so they reinvent all of it.
Karbon assumes Tophan. It knows the Networking Scale handles pod networking. It knows the Storage Scale handles volumes. It knows the Security Scale handles policy. This means Karbon can be simpler, faster, and more reliable — because it’s built on known foundations instead of abstraction layers.
┌──────────────────────────────┐
│ Dragon's Eye │
├──────────────────────────────┤
│ Karbon Scale API │ Scheduling, lifecycle
├──────┬──────┬────────────────┤
│ Net │ Stor │ Layers Scale │ Integrated Scales
├──────┴──────┴────────────────┤
│ tophand │ Node orchestration
└──────────────────────────────┘
Karbon sits on top of the Layers Scale (for workload execution), the Networking Scale (for connectivity), and the Storage Scale (for persistence). It adds scheduling, scaling, and lifecycle management without reimplementing what those Scales already provide.