Mr.PlanB Logo
    On-Prem Kubernetes Ledger 2026
    Ledger
    On-Prem Kubernetes Ledger 2026

    DaoCloud Enterprise: On-Prem Kubernetes Ledger

    DaoCloud sells DCE as a set of independently versioned modules (Karmada multicloud, KubeVirt VMs, Harbor registry, Argo CD workbench, OpenTelemetry…

    Edition October 13, 20266 min read

    Part of the On-Prem Kubernetes Ledger (edition 2026-10-13). How the scores work: methodology.

    1. Snapshot

    • Tier: Regional
    • Owner: Shanghai DaoCloud Network Technology Co., Ltd., privately held, venture funded (Series D reported 2024-02-01)
    • Licence model: Commercial subscription, no public price list; trial and downloads on request; open-source projects Kubean, HwameiStor, Spiderpool, Clusterpedia under Apache 2.0
    • Products scored: DCE 5.0 modules: container management, multicloud, global management, workbench, insight, registry, mesh, VirtNest VMs, AI Lab and LLM Studio
    • Latest release: Installer v0.44.0 (2026-08-31, default Kubernetes 1.35.5 from v0.42.0); CNCF certified for Kubernetes 1.36 on 2026-08-14
    • Next signal: No roadmap published
    • Archetype: Modular China platform
    • Evidence grade: B. Release notes, docs repository and CNCF repositories were opened; the English docs are thin on support windows, FIPS, DRA, MCP and migration tooling, and the Chinese docs were not read.

    2. Scorecard

    DimensionScoreDirection
    Capability today72.2 / 100higher is better
    Momentum since end of 2022+27.0 pointshigher is better
    Credibility17.0 / 25higher is better
    Commercial risk13.5 / 25lower is better
    Technical lock-in10.0 / 25lower is better
    Integrator fit11.5 / 20higher is better
    AI leverage8.5 / 20higher is better
    Ledger Index64.5 / 100balanced weights

    3. Capability by domain

    Scores 0 to 5 against fixed anchors; total is weighted to 100.

    Domain (weight)201620192022Today
    Install, upgrade and lifecycle (12)n/an/a2.53
    Bare metal, edge and on-prem infrastructure (10)n/an/a2.53.5
    Multi-cluster and fleet management (10)n/an/a34
    Networking (10)n/an/a34
    Storage and data protection (10)n/an/a2.53.5
    Security and compliance (12)n/an/a2.53
    Developer platform (8)n/an/a3.54
    Observability and day-2 operations (8)n/an/a3.54
    VMs on Kubernetes (VMware exit) (10)n/an/a03.5
    GPU and accelerator workloads (10)n/an/a04
    Total / 100not releasednot released45.272.2

    Year by year: 2022: 45.2 · 2023: 53.4 · 2024: 68.2 · 2025: 71.2 · 2026: 72.2

    What moved the score

    YearDomainChangePointsTrigger
    2023Bare metal, edge and on-prem infrastructure2.5 → 3+1.0Cloud-edge module v0.5.0 (2023-11); OS list across amd64 and arm64 (Red Hat, Ubuntu, Rocky, openEuler, UOS, Kylin)
    2023Multi-cluster and fleet management3 → 3.5+1.0Multicloud v0.12.0 (2023-08): cross-cluster horizontal autoscaling; propagation policies; DCE4 to DCE5 app migration
    2023Security and compliance2.5 → 3+1.2CIS compliance, authorization and vulnerability scans, LDAP, OIDC and OAuth2 login, audit log, Falco guide; no FIPS or SBOM
    2023Storage and data protection2.5 → 3+1.0HwameiStor v0.13.1 (2023-10): LVM snapshots; DRBD replication; Velero-based backup module
    2023VMs on Kubernetes (VMware exit)0 → 2+4.0VirtNest v0.1.0 (2023-07) and v0.2.0 (2023-10): create, console, snapshots, clone, restore
    2024Developer platform3.5 → 4+0.8Workbench v0.31.0 (2024-09): ApplicationSet, Argo CD UI, canary and blue-green; middleware catalogue; no developer portal
    2024GPU and accelerator workloads0 → 3+6.0Container management v0.32.0 (2024-09): Volcano binpack and priority; NVIDIA GPU Operator and HAMi vGPU docs; Cambricon cards (2024-10)
    2024Install, upgrade and lifecycle2.5 → 3+1.2Monthly installer, UI cluster upgrades, default Kubernetes raised yearly to 1.35.5 (2026-06); median lag 142 days; no rollback found
    2024Multi-cluster and fleet management3.5 → 4+1.0Workbench v0.32.0 (2024-10): GitOps into multicloud namespaces; no cluster-count limit or fleet compliance feature found
    2024Networking3 → 4+2.0SR-IOV node policies and RDMA (2024-01), Multus, EgressGateway, Ambient mesh injection (2024-09); no IPv6 or Gateway API statement
    2024Observability and day-2 operations3.5 → 4+0.8Insight v0.31 (2024-09): topology maps, log alerts; GPU dashboards and slow SQL analysis later; no cost reporting
    2024VMs on Kubernetes (VMware exit)2 → 3.5+3.0Live migration, scheduled snapshots, templates, Windows guests, vGPU, vSphere import guide (2024-07); latest v0.18.1 (2025-07-30); no bulk toolkit
    2025Bare metal, edge and on-prem infrastructure3 → 3.5+1.0Device Management v0.4.1 (2025-08): Redfish onboarding and PXE provisioning; no hypervisor providers
    2025GPU and accelerator workloads3 → 4+2.0GPU management v0.41.0 (2025-07), MetaX cards, third-party device plugins, MIG; AI Conformance 2025-10-31; no DRA feature found
    2026Storage and data protection3 → 3.5+1.0etcd backup to HTTPS object storage (v0.49.0, 2026-07); no array-vendor certification list found

    4. Direction, 2023 to 2026

    Each item carries one theme and one driver (V vision, C customer need, M market window, U upstream, P portfolio).

    DateItemThemeDriver
    2023-07VirtNest v0.1.0: virtual machines on KubeVirt in the consoleD10 VMs on KubernetesM
    2023-08Multicloud v0.12.0: cross-cluster horizontal autoscaling (FHPA)D3 Fleet managementV
    2023-11Cloud-edge collaboration module v0.5.0D2 Bare metal, edge and infrastructureM
    2024-04AI Lab v0.3.0: notebooks, training jobs, model serving, data managementD13 AcceleratorsM
    2024-09Container management v0.32.0: Volcano binpack and priority scheduling for GPU jobsD13 AcceleratorsC
    2024-10Workbench v0.32.0: GitOps applications into multicloud namespacesD3 Fleet managementC
    2025-07GPU management in container management; LLM Studio v0.7.0D13 AcceleratorsM
    2025-08Device Management v0.4.1: Redfish onboarding and PXE provisioningD2 Bare metal, edge and infrastructureC
    2025-10CNCF Kubernetes AI Conformance passed for Kubernetes 1.33D13 AcceleratorsM
    2026-03ClawOS v0.2.0: managed AI agent runtime on DCED6 Developer platformM
    2026-06Installer default Kubernetes raised to 1.35.5 (v0.42.0)D1 Install and lifecycleU
    2026-08Installer v0.44.0: basic GitOps-mode deployment frameworkD1 Install and lifecycleV
    • Centre of gravity: Accelerators (4), Fleet management (2), Bare metal, edge and infrastructure (2)
    • Driver mix: Vision 17%, Customer need 25%, Market window 50%, Upstream 8%, Portfolio 0%

    5. Credibility: 17.0 / 25

    CriterionScoreBasis
    Cadence4.5Month-end release notes per module; 48 installer versions from 2022-11-30 to 2026-08-31 without a gap in the last 24 months
    Roadmap transparency3Public docs repository with release notes and open-source projects; no roadmap or milestones
    Say-do2.5No dated public promises found, so nothing to test; one limitation (default Kubernetes 1.24) cleared in about 7 months
    Velocity4Many capabilities per year: VMs, AI Lab, LLM Studio, GPU management, Device Management, ClawOS; VirtNest quiet since 2025-07
    Lifecycle stability3No support windows published; DCE 4 to 5 needed a limited-scenario migration; product lines split into d.run lines 2026

    Say-do record, 2023 to 2026: 0 dated promises found 2023 to 2026; nothing counted. One known limitation (default Kubernetes stuck at 1.24 in 2022-11) cleared by 2023-06-30.

    6. Commercial risk: 13.5 / 25 (lower is better)

    CriterionScoreBasis
    Price and licence volatility2No licence or price change found; product lines reorganised into d.run Token Factory, d.run AI OS and DCE
    Purchase constraints3Quote-based subscription, free trial and download pages; no public price or minimum
    Channel and access2.5Docs and release notes open; English coverage thin; ecosystem page, partner terms not read
    Owner stability3Private, venture funded; Series D reported 2024-02-01; no profitability data found
    Cost of staying supported3No published support window or extended support; monthly updates; some modules paid

    Security record of the product itself: No KEV entry. DaoCloud-sponsored components: Kubean CVE-2024-41820 (6.0, fixed 0.18.0 on 2024-08-05, same day as advisory), HwameiStor CVE-2024-36534 (8.4) and CVE-2024-45054 (2.8, fixed 0.14.6 on 2024-07-29), Spiderpool CVE-2024-33393 (6.2). No critical CVE. No DCE advisory page found. From security-record corrections.

    7. Technical lock-in: 10.0 / 25 (lower is better)

    CriterionScoreBasis
    Formats1.5CNCF certified for Kubernetes 1.24 to 1.36; lifecycle through open-source Kubean and Karmada; standard workload APIs
    Export path2.5DCE4 to DCE5 migration guide and Velero backups; no guide for leaving DCE
    Stack coupling2CNI and storage choices (Cilium, Calico, Spiderpool, HwameiStor, any CSI); broad OS list
    Hardware / cloud coupling1.5Any amd64 or arm64 hardware, offline install; GPU and Cilium features need kernel 5.12 or later
    Skills and tooling coupling2.5DCE module ecosystem and workspace model; documentation Chinese first; Kubespray-style lifecycle

    8. Integrator fit: 11.5 / 20

    CriterionScoreBasis
    Partner programme openness2.5Ecosystem page in docs hub; partner terms not read
    Multi-tenancy and self-service4Global Management workspaces, resource groups, custom roles, LDAP, OIDC and OAuth2 login
    API and infrastructure-as-code3Kubean, Helm app store, module OpenAPI; no Terraform or Cluster API provider found
    Skills and certification2No certification found; open-source community around Kubean, HwameiStor and Spiderpool

    9. AI leverage: 8.5 / 20

    StrandTodayLatest step
    AI workload support32025: CNCF Kubernetes AI Conformance passed 2025-10-31; vLLM inference services in AI Lab; dataset spaces
    AI platform stack42025: LLM Studio v0.7.0 and v0.8.0 (2025-07, 2025-08): model catalogue, OpenAI-compatible API, rate limits, vLLM deployment
    AI-assisted operations0none
    Agent openness1.52022: Documented OpenAPI per module and Kubernetes API; ClawOS (2026-03) runs customer agents but is not an MCP server

    AI say-do: 0 promises found; 5 AI items delivered (AI Lab, LLM Studio, AI Conformance, ClawOS, Volcano); no DCE operator assistant and no DCE MCP server found.

    10. Reading

    DaoCloud sells DCE as a set of independently versioned modules (Karmada multicloud, KubeVirt VMs, Harbor registry, Argo CD workbench, OpenTelemetry observability) and has moved toward GPU, AI Lab and LLM Studio since 2024. The record is steady monthly delivery and an unbroken CNCF certification line for 13 minors, but no public roadmap, no stated promises to test and a Kubernetes lag of about 142 days. A buyer gets broad coverage and open-source components; the English documentation is thin on support terms and compliance.

    11. Open questions and evidence caveats

    • English docs only: FIPS, DRA, MCP, support windows and VMware migration tooling may exist in Chinese docs or sales material and are scored as not found.
    • The first-shipped date per Kubernetes minor is a proxy (earlier of CNCF certification and installer default).
    • VirtNest has no release since 2025-07-30 in the docs; the VM score assumes it is still sold and supported.
    • No pricing, no support-window policy and no Terraform or Cluster API provider were found.

    More from On-Prem Kubernetes Ledger 2026