Scody Cloud

Infrastructure

We do not resell someone else's control panel. The platform is ours — from the hypervisor to the metering pipeline.

Compute

Proxmox VE clusters running KVM guests and LXC containers, provisioned declaratively with Terraform and configured with Ansible.

Storage

Local NVMe with ZFS at the edge, triple-replicated Ceph across the cluster, and an S3-compatible RGW endpoint for object storage.

Network

VLAN and VXLAN tenant isolation, VyOS edge routing, Netbox as the source of truth for every IP, prefix and host.

Control plane

An in-house provisioning API that creates, resizes, rebuilds and destroys resources, plus a metering pipeline for CPU, RAM, disk and transfer.

Operations

Prometheus, Grafana and Alertmanager for metrics, Loki for logs, Restic to offsite object storage, with restore drills before anything ships.

Build roadmap

  1. Stage 1

    Foundation

    Two bare-metal nodes, Proxmox cluster, Ansible baseline, monitoring and verified backups.

  2. Stage 2

    Control plane

    Provisioning API, Packer image pipeline, IPAM, tenant isolation and usage metering.

  3. Stage 3

    Storage & scale-out

    Third node, Ceph migration, block and object storage, automated capacity thresholds.

  4. Stage 4

    Portal & billing

    Customer dashboard, catalogue, ordering, invoices, API keys and payments.

  5. Stage 5

    Dedicated & cloud

    Bare-metal automation via IPMI/Redfish, hourly cloud flavors, snapshots and load balancing.

  6. Stage 6

    Serverless GPU

    Kubernetes with the NVIDIA GPU Operator, queue-driven autoscaling and per-second billing.