Skip to content

Mirantis k0rdent Enterprise v1.2.3 Release Notes#

Released: January 29, 2026

Mirantis k0rdent Enterprise builds on the upstream, community-driven k0rdent OSS project to provide a commercially supported, enterprise-grade environment for managing Kubernetes clusters, services, and observability. While the open source k0rdent delivers core functionality under the Apache 2.0 license, Mirantis k0rdent Enterprise adds hardened components, tested integrations, and enterprise-only features—including a fully-featured UI, the ability to add a custom certificate authority, and bare metal provisioning.

Component & Provider Versions#

Component / Provider Version
Cluster API v1.11.2
CAPI Provider AWS v2.10.0
CAPI Provider Azure v1.21.0
CAPI Provider Docker v1.11.2
CAPI Provider GCP v1.10.0
CAPI Provider Infoblox v0.1.0
CAPI Provider IPAM v1.1.0-rc.1
CAPI Provider k0smotron v1.10.0
CAPI Provider OpenStack (ORC) v0.12.5-mirantis.0 / v2.1.0
CAPI Provider vSphere v1.14.0
Project Sveltos v1.1.1

Grafana not Included#

Effective immediately, Mirantis will no longer distribute Grafana as part of its products or services. This change is being made to proactively avoid potential licensing, redistribution, or compliance considerations related to third-party software.

Grafana dashboards and grafana-operator (for installation and lifecycle management of Grafana on the k0rdent Enterprise management cluster or on child clusters) will still be shipped as part of k0rdent Observability & FinOps (KOF), an optional component of k0rdent Enterprise. For more information, please contact Mirantis.

For instructions on how to install Grafana locally and integrate with Mirantis Grafana dashboards and grafana-operator, please see Grafana in KOF.

Highlights#

  • ARM64 and Multi-Cluster Support Mirantis k0rdent Enterprise now runs on arm64 architectures and manages larger multi-cluster fleets without separate tooling. This removes hardware constraints and lets operators apply one consistent platform across datacenter, edge, and mixed-environment footprints.

  • Enterprise Identity & Authorization Management A unified identity layer now ships with Mirantis k0rdent Enterprise, offering built-in authentication, federation with enterprise IdPs, and the option to plug in a fully external identity system. This creates a single, scalable model for user access across management and child clusters without locking operators into one approach.

  • External Database Support for Hosted Control Planes Hosted control planes can now use Kine backed by an external relational database. Offloading state to a dedicated database increases resilience, smooths performance under load, and supports higher cluster density than embedded etcd permits.

  • Distributed Regional Management Control Planes Hosted control plane pods can be deployed on designated regional Kubernetes clusters instead of the primary management cluster. This adds clean separation between management and regional domains, improves failure isolation, and allows credential boundaries to match organizational or geographic requirements.

  • k0rdent Cluster Manager (KCM) Enhancements KCM now includes telemetry collection, configurable Helm operation timeouts, support for regional HCP placement, and label-based role aggregation. These updates reduce friction during upgrades, simplify RBAC integration, and make lifecycle automation more predictable at scale.

  • ServiceSet API & KSM Enhancements The new ServiceSet abstraction allows operators to define services with explicit dependencies and ordered upgrades, eliminating ad-hoc sequencing. Additional Helm options, dependency awareness, and improved IPAM observability make multi-component rollouts far more controlled and transparent.

  • Observability (KOF) Improvements KOF adds kube-state-metrics dashboards for k0rdent CRDs, a raw-metrics viewer, Promxy tuning for heavy queries, and monitoring for KSM objects. The result is clearer insight into cluster behavior and fewer blind spots when diagnosing deployment or reconciliation issues.

  • Regional and Fleet-Wide Operational Improvements Regions can now be restored from backup, clusters can migrate between management clusters, identity data can propagate across regions, and a dedicated RBAC manager standardizes permissions. Operators can also pause Sveltos reconciliation, giving them safer control during troubleshooting and disaster recovery workflows.

  • Access Management Resource The new AccessManagement resource provides a centralized, rule-based mechanism for distributing ClusterTemplates, ServiceTemplates, Credentials, and ClusterAuthentication objects across namespaces. It removes the need for manual object propagation and ensures that teams or tenants receive only the resources intended for them. This brings consistency to multi-namespace environments and reduces the risk of configuration drift when managing large numbers of clusters.

  • Bare Metal Management from the UI In addition to the command line, you can now manage Bare Metal machines and ClusterDeployments from the UI, enabling unified management from a single platform.

  • UI Upgrade The k0rdent UI has been upgraded to version 1.1.1 with numerous fixes and improvements.

  • Stability & UX: Multiple fixes to credential propagation, service status collection, hosted cluster templates, Helm repos, and cloud provider templates (OpenStack, Azure).

🚀 New Features 🚀#

🐛 Notable Fixes 🐛#

✨ Notable Changes ✨#

  • fix(os-tpls): correct identity name in identityref (#1901) by @zerospiel
  • refactor: reuse already defined statemanagementprovider name (#1883) by @wahabmk
  • test: check promxy metrics by @gmlexx
  • test: wait until vmauth creates ingress in kind-adopted-regional cluster (#471) by @gmlexx
  • test: add unit tests for Victoria pages (KOF UI) (#482) by @AndrejsPon00

❗ Upgrade Notes ❗#

  • If you have non-airgapped k0rdent cluster please apply the following steps before upgrading:

    1. Get current k0sURL used in the system, by executing for any of cluster deployments you have:

      kubectl -n kcm-system get hr <cluster deployment name> -o jsonpath='{.spec.values.global.k0sURL}'

      Note

      If the returned value is https://get.mirantis.com/k0rdent-enterprise/k0s you may skip all next steps and proceed with the upgrade.

    2. Update the Management object by putting the value from the step a. to the .spec.core.kcm.config.controller.globalK0sURL parameter.

    3. Proceed with the upgrade as normal

  • If upgrade in an airgapped environment is stuck with "waiting for capi" conditions in Management and theCoreProvider object has the config map not found condition, you have hit the known issue kubernetes-sigs/cluster-api-operator#966. To solve this problem, execute the following steps:

    1. Download and re-upload the cluster-api-provider-k0sproject-k0smotron-components image (replace REGISTRY with the airgap registry hostname):

      skopeo copy -a --insecure-policy docker://registry.mirantis.com/k0rdent-enterprise/capi/cluster-api-provider-k0sproject-k0smotron-components:v1.6.0 oci-archive:cluster-api-provider-k0sproject-k0smotron-components_v1.6.0.tar
skopeo copy -a --insecure-policy oci-archive:cluster-api-provider-k0sproject-k0smotron-components_v1.6.0.tar docker://${REGISTRY}/k0rdent-enterprise/capi/cluster-api-provider-k0sproject-k0smotron-components:v1.6.0

    2. Delete the capi-operator pod to restart the reconcile process.

  • New/stricter Region validation may surface issues with configurations that were previously accepted. Dry-run manifests before rollout.

  • Region restoration flow is new; confirm backup/restore procedures for regional components.

  • CI images switched base images from scratch to gcr.io/distroless/static-debian12:nonroot

Known Issues#

  • Unexpected child cluster node rotation after upgrade (k0rdent-enterprise#159) caused by k0sURL update. Only non-airgap clusters are affected, please follow the upgrade notes to mitigate the issue.
  • kof-mothership chart may fail on removal in case if specific mcs are present. To avoid this, please delete all corresponding MultiClusterServices before uninstalling kof-mothership chart (kubectl delete mcs kof-storage-secrets kof-storage-secrets-remote-templates-copy).
  • k0rdent-istio requires two workarounds documented in Installing KOF - Istio section.
  • KOF UI shows false positive misconfiguration warnings, fixed in the next release.
  • Upgrade can get stuck indefinitely due to CAPI not upgrading, causing all components to show the "waiting for capi" condition. This is related to the upstream issue kubernetes-sigs/cluster-api-operator#966. Check upgrade notes for mitigation steps.

Release Metadata#

Key Value
Helm Charts kcm: 1.2.2, kof: 1.5.0
OCI Registry registry.mirantis.com/k0rdent-enterprise/
SBOM Included
OCI Signature Support Included
Release Tags v1.2.2 across all components

Contributors#

Huge thanks to the following contributors for making this release possible: @gmlexx, @denis-ryzhkov, @aglarendil, @kylewuolle, @a13x5, @eromanova, @zerospiel, @BROngineer, @Kshatrix, @dis-xcom, @wahabmk, @AndrejsPon00

Resources#

Try It Out#

QuickStart guide: https://docs.k0rdent-enterprise.io/v1.2.2/quickstarts/