kranix-charts

Helm charts — install the full Kranix platform on Kubernetes in one command.

kranix-charts contains the official Helm charts for deploying the Kranix platform to a Kubernetes cluster. It packages kranix-api, kranix-core, kranix-operator, kranix-mcp, and their dependencies into a single, configurable release. This is the recommended way to run Kranix in any environment beyond local development.

What it does

Packages all Kranix components into a single Helm release
Manages CRD installation alongside the application charts
Provides sane defaults with full configurability via values.yaml
Supports multi-environment patterns (dev, staging, production overlays)
Includes RBAC, ServiceAccounts, NetworkPolicies, and PodDisruptionBudgets
Supports optional components (kranix-mcp, metrics, ingress) via feature flags
Production-ready profile with HA replicas, PodDisruptionBudgets, and anti-affinity rules
OpenTelemetry integration for distributed tracing and metrics
Pre-upgrade validation hooks to prevent failed upgrades
Standalone chart installation for individual components
ArgoCD app-of-apps template for GitOps deployment
Resource sizing calculator for automatic resource recommendations

Architecture position

Helm CLI  ──►  kranix-charts
                    │
                    ├── kranix-core (Deployment)
                    ├── kranix-api (Deployment + Service)
                    ├── kranix-operator (Deployment)
                    ├── kranix-mcp (Deployment + Service) [optional]
                    ├── CRDs (KranixApp, KranixNamespace, KranixPolicy)
                    └── RBAC (ClusterRole, ClusterRoleBinding, ServiceAccounts)

Chart structure

kranix-charts/
├── charts/
│   ├── kranix/                  # Umbrella chart (installs everything)
│   │   ├── Chart.yaml
│   │   ├── values.yaml         # Default values
│   │   ├── templates/
│   │   │   ├── _helpers.tpl
│   │   │   ├── namespace.yaml
│   │   │   └── crds/           # CRD templates
│   │   └── charts/             # Sub-charts (vendored)
│   ├── kranix-core/             # Core engine chart
│   ├── kranix-api/              # API server chart
│   ├── kranix-operator/         # Operator chart
│   └── kranix-mcp/              # MCP server chart (optional)
├── ci/                         # CI values for chart testing
└── docs/                       # values.yaml reference docs

Prerequisites

Kubernetes 1.27+
Helm 3.12+
kubectl configured for your target cluster

Install

Add the Kranix Helm repository

helm repo add kranix https://charts.kranix.io
helm repo update

Install full platform (umbrella chart)

helm install kranix kranix/kranix \
  --namespace kranix-system \
  --create-namespace

Install individual components (standalone charts)

Each component can be installed independently:

# Install only kranix-core
helm install kranix-core kranix/kranix-core \
  --namespace kranix-system \
  --create-namespace

# Install only kranix-operator
helm install kranix-operator kranix/kranix-operator \
  --namespace kranix-system \
  --create-namespace

# Install only kranix-mcp
helm install kranix-mcp kranix/kranix-mcp \
  --namespace kranix-system \
  --create-namespace

Install with custom values

helm install kranix kranix/kranix \
  --namespace kranix-system \
  --create-namespace \
  --values ./my-values.yaml

Verify installation

kubectl get pods -n kranix-system
kubectl get crds | grep kranix.io

Configuration (`values.yaml`)

Global

global:
  image:
    registry: ghcr.io/kranix-io
    pullPolicy: IfNotPresent
  serviceAccount:
    create: true
    name: kranix

kranix-core

core:
  enabled: true
  replicas: 1
  image:
    tag: "latest"
  config:
    reconcile_interval: 15s
    max_concurrent_reconciles: 10
  resources:
    requests:
      cpu: 100m
      memory: 128Mi
    limits:
      cpu: 500m
      memory: 512Mi

kranix-api

api:
  enabled: true
  replicas: 2
  image:
    tag: "latest"
  service:
    type: ClusterIP
    port: 8080
  ingress:
    enabled: false
    className: nginx
    host: kranix.example.com
    tls: true
  auth:
    mode: jwt                  # jwt | apikey | oidc
    jwtSecret: ""              # use secretRef in production
  resources:
    requests:
      cpu: 100m
      memory: 128Mi
    limits:
      cpu: 500m
      memory: 256Mi

kranix-operator

operator:
  enabled: true
  replicas: 1
  image:
    tag: "latest"
  leaderElection: true
  resources:
    requests:
      cpu: 50m
      memory: 64Mi
    limits:
      cpu: 200m
      memory: 128Mi

kranix-mcp (optional)

mcp:
  enabled: false
  replicas: 1
  image:
    tag: "latest"
  service:
    type: ClusterIP
    port: 3100
  safety:
    readonlyMode: false
    allowDeleteWorkload: true

State backend

state:
  backend: postgres            # memory | postgres | etcd
  postgres:
    host: ""
    port: 5432
    database: kranix
    existingSecret: kranix-postgres-secret
    secretKey: postgres-password

Metrics and observability

metrics:
  enabled: true
  port: 9090
  serviceMonitor:
    enabled: false             # set true if using kube-prometheus-stack

tracing:
  enabled: false
  endpoint: ""                 # OTLP endpoint

logging:
  level: info
  format: json

Production-ready profile

Enable high-availability configurations with a single flag:

kranix-core:
  productionReady:
    enabled: true
    replicas: 3
    podDisruptionBudget:
      enabled: true
      minAvailable: 2
    antiAffinity:
      enabled: true
      type: hard                # hard or soft

When enabled, this profile:

Sets high replica counts for HA
Creates PodDisruptionBudgets to ensure minimum availability during disruptions
Configures pod anti-affinity rules to spread pods across nodes
Can be configured per-component (kranix-core, kranix-api, kranix-operator)

OpenTelemetry integration

Enable distributed tracing and metrics collection:

kranix-core:
  opentelemetry:
    enabled: true
    endpoint: "http://opentelemetry-collector.observability.svc.cluster.local:4317"
    tracing:
      enabled: true
      samplingRatio: 1.0
    metrics:
      enabled: true
      port: 9464
      path: /metrics
    resourceAttributes:
      service.name: kranix-core
      deployment.environment: production

Features:

Exports traces to OpenTelemetry Collector via OTLP
Exposes Prometheus metrics endpoint for scraping
Configurable sampling ratio for traces
Service-level resource attributes for filtering

Pre-upgrade validation hooks

Run validation checks before Helm upgrades to prevent failures:

kranix-core:
  preUpgradeHook:
    enabled: true
    image:
      repository: bitnami/kubectl
      tag: "latest"
    checks:
      resources:
        enabled: true
        minCpu: "100m"
        minMemory: "128Mi"
      dependencies:
        enabled: true
        services:
          - postgres
      secrets:
        enabled: true
        required:
          - kranix-postgres-secret

Validation checks include:

Resource availability (CPU, memory)
Dependent service readiness
Required secret existence
Hook runs as a Kubernetes Job with pre-upgrade annotation

Resource sizing calculator

Automatically calculate recommended CPU and memory resources based on workload count and deployment tier:

kranix-core:
  resourceCalculator:
    enabled: true
    workloadCount: 500           # Expected number of workloads
    tier: medium                 # small, medium, or large

The calculator provides tier-based resource recommendations:

Small tier (up to 200 workloads):

kranix-core: 100m-500m CPU, 128Mi-512Mi memory
kranix-api: 100m-500m CPU, 128Mi-512Mi memory
kranix-operator: 50m-200m CPU, 64Mi-256Mi memory

Medium tier (up to 500 workloads):

kranix-core: 200m-1000m CPU, 256Mi-1Gi memory
kranix-api: 200m-1000m CPU, 256Mi-1Gi memory
kranix-operator: 100m-500m CPU, 128Mi-512Mi memory

Large tier (2000+ workloads):

kranix-core: 1000m-4000m CPU, 512Mi-4Gi memory
kranix-api: 1000m-4000m CPU, 512Mi-4Gi memory
kranix-operator: 500m-2000m CPU, 256Mi-2Gi memory

When enabled, the calculator overrides manual resource settings with calculated values.

ArgoCD GitOps deployment

The charts include an ArgoCD Application manifest for GitOps-based deployment:

# Add Helm repository to ArgoCD
argocd repo add https://charts.kranix.io --type helm --name kranix

# Deploy via ArgoCD
kubectl apply -f argocd/kranix-platform-app.yaml

See argocd/README.md for detailed GitOps deployment instructions.

Multi-environment patterns

Using Helm values overlays

# Base install
helm install kranix kranix/kranix -f values-base.yaml

# Environment-specific overlay
helm upgrade kranix kranix/kranix \
  -f values-base.yaml \
  -f values-production.yaml

Example production overlay (`values-production.yaml`)

api:
  replicas: 3
  ingress:
    enabled: true
    host: kranix.mycompany.com
    tls: true
  auth:
    mode: oidc
  # Enable production-ready profile
  productionReady:
    enabled: true
    replicas: 3
  # Enable OpenTelemetry
  opentelemetry:
    enabled: true
  # Enable pre-upgrade validation
  preUpgradeHook:
    enabled: true
  # Enable resource calculator
  resourceCalculator:
    enabled: true
    workloadCount: 500
    tier: medium

mcp:
  enabled: true

state:
  backend: postgres

metrics:
  serviceMonitor:
    enabled: true

Upgrade

helm repo update
helm upgrade kranix kranix/kranix \
  --namespace kranix-system \
  --reuse-values

Check the CHANGELOG.md before upgrading across major versions — some releases include CRD schema changes.

Rollback

helm rollback kranix 1 --namespace kranix-system

Uninstall

helm uninstall kranix --namespace kranix-system

# Remove CRDs (warning: deletes all KranixApp resources)
kubectl delete crds \
  kraneapps.kranix.io \
  kranenamespaces.kranix.io \
  kranepolicies.kranix.io

Connectivity

Repo	Relationship
`kranix-core`	Packaged as a sub-chart and deployed as a Deployment
`kranix-api`	Packaged as a sub-chart with a Service and optional Ingress
`kranix-operator`	Packaged as a sub-chart with CRDs and RBAC
`kranix-mcp`	Optional sub-chart, disabled by default

Contributing

See CONTRIBUTING.md. Use helm lint and chart-testing (ct) for all chart changes. CI runs ct install against a kind cluster on every PR.

ct lint --charts charts/kranix
ct install --charts charts/kranix

License

Apache 2.0 — see LICENSE.

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
argocd		argocd
charts		charts
.gitignore		.gitignore
CHANGELOG.md		CHANGELOG.md
CONTRIBUTING.md		CONTRIBUTING.md
LICENSE		LICENSE
README.md		README.md
ct.yaml		ct.yaml

Folders and files

Latest commit

History

Repository files navigation

kranix-charts

What it does

Architecture position

Chart structure

Prerequisites

Install

Add the Kranix Helm repository

Install full platform (umbrella chart)

Install individual components (standalone charts)

Install with custom values

Verify installation

Configuration (values.yaml)

Global

kranix-core

kranix-api

kranix-operator

kranix-mcp (optional)

State backend

Metrics and observability

Production-ready profile

OpenTelemetry integration

Pre-upgrade validation hooks

Resource sizing calculator

ArgoCD GitOps deployment

Multi-environment patterns

Using Helm values overlays

Example production overlay (values-production.yaml)

Upgrade

Rollback

Uninstall

Connectivity

Contributing

License

About

Resources

License

Contributing

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Configuration (`values.yaml`)

Example production overlay (`values-production.yaml`)

Packages