OpenSDDC ecosystem & projects

Shared logical thin volume model with a QEMU-aware LVM approach.

SLTV does not aim to copy VMFS verbatim; it targets an operational experience on KVM/QEMU over shared storage similar to what teams expect from enterprise stacks. The goal is to keep existing SAN investments while enabling thin provisioning, live migration, and VM lifecycle flows on KVM platforms.

• KVM/QEMU volume model on shared LUN
• QEMU-aware LVM approach
• Thin provisioning behavior
• Live migration
• SAN, iSCSI, FC and future CSI integration

Management, metrics and telemetry with eBPF over gRPC.

gIMI targets reducing the sprawl of exporters, agents, vendor APIs, SNMP, syslog, and custom collectors used to pull data from infrastructure. The goal is to observe host, VM, Kubernetes node, workload, network, and storage in a common event chain.

• gRPC-based infrastructure management API
• Shared API surface for management, metrics, and telemetry
• eBPF support
• OpenTelemetry-compatible export target

Opinionated datacenter KVM Linux.

dc(e)OS is not aiming to be a new general-purpose Linux distribution. The goal is a manageable, consistent host OS profile for datacenter nodes with KVM/QEMU, libvirt, OVS/OVN, SLTV, and gIMI integrations.

• RHEL / Rocky / Alma-aligned baseline approach
• Immutable host OS
• KVM / QEMU / libvirt-ready profile
• Podman / container runtime-ready profile
• OVS / OVN-ready profile
• Node model manageable by OpenSDDC Manager

Inventory, service map, incident timeline, and automation layer.

OpenDIMA represents the inventory, service map, metric/log correlation, incident context, and automation layer in the OpenSDDC ecosystem. The goal is to make sense of Kubernetes, virtualization, network, storage, and security in one event chain.

• Discovery and inventory
• Service map
• Telemetry support via gIMI
• Incident timeline
• Alertmanager-style alert flows

Datacenter management platform for dc(e)OS clusters and more.

OpenSDDC Manager aims to unify dc(e)OS clusters, multi-hypervisor environments (Proxmox, vCenter, Hyper-V, cloud, etc.), and Kubernetes clusters with OpenDIMA event context in a single control plane.

• Multi-hypervisor: dc(e)OS, Proxmox, ESXi
• Unified inventory and view
• Pod ↔ VM ↔ host event chain
• gIMI telemetry correlation
• Plugin / adapter model (Proxmox, vCenter, Kubernetes)
• Policy-driven workload placement target