Browse docs
--- title: "Kubernetes and Helm" description: "Deploy Aurora on Kubernetes using the Helm chart with support for OSS and Enterprise configurations." icon: "layers" --- ## Kubernetes and Helm The Aurora Helm chart provides a production-grade Kubernetes deployment with configurable providers, persistence, caching, monitoring, and Enterprise cluster support. ### Prerequisites - Kubernetes 1.29+ - Helm 3.x - (Optional) Prometheus Operator for ServiceMonitor support ### Quick Install with Default Values ``bash # Add the Helm repository (if published) # helm repo add aurora https://aurorallm.github.io/Aurora/charts # Install from local chart directory helm install aurora ./helm \ -n aurora --create-namespace \ --set providers.openai.apiKey="sk-..." ` ### Multi-Provider Setup `bash helm install aurora ./helm \ -n aurora --create-namespace \ --set providers.openai.apiKey="sk-..." \ --set providers.anthropic.apiKey="sk-ant-..." ` ### Using Existing Secrets (GitOps) `bash helm install aurora ./helm \ -n aurora --create-namespace \ --set providers.existingSecret="llm-api-keys" \ --set providers.openai.enabled=true \ --set providers.anthropic.enabled=true ` ### Enterprise Deployment For Enterprise deployments, add license key and configure enterprise features: `bash helm install aurora ./helm \ -n aurora --create-namespace \ --set providers.existingSecret="llm-api-keys" \ --set auth.masterKey="your-master-key" \ --set redis.enabled=true \ --set metrics.enabled=true ` Enterprise license keys, private container images, and deployment profiles are provided during onboarding. Contact Aurora for access. ## Key Configuration Values | Parameter | Default | Description | |---|---|---| | replicaCount | 2 | Number of replicas | | server.port | 8080 | Server port | | auth.masterKey | "" | Admin authentication key | | auth.existingSecret | "" | Existing secret for master key | | providers.existingSecret | "" | Existing secret for all provider API keys | | redis.enabled | true | Deploy Redis subchart (Bitnami Redis) | | redis.architecture | standalone | Redis architecture: standalone or replication | | metrics.enabled | true | Enable Prometheus metrics | | metrics.serviceMonitor.enabled | false | Create ServiceMonitor for Prometheus Operator | | ingress.enabled | false | Enable Kubernetes Ingress | | gateway.enabled | false | Enable Gateway API HTTPRoute | | autoscaling.enabled | false | Enable HPA (minReplicas: 2, maxReplicas: 10) | | podDisruptionBudget.enabled | true | Enable PDB (minAvailable: 1) | | resources.requests.cpu | 100m | CPU request per pod | | resources.requests.memory | 128Mi | Memory request per pod | ## Production Defaults The Helm chart ships with production-friendly defaults: - 2 replicas by default for HA - PodDisruptionBudget enabled (min 1 available) - Health probes configured on /health - Security context — read-only root filesystem, non-root user (UID 1000), all capabilities dropped - Redis deployed as a subchart for caching ### Security Context `yaml podSecurityContext: runAsNonRoot: true runAsUser: 1000 fsGroup: 1000 securityContext: allowPrivilegeEscalation: false readOnlyRootFilesystem: true runAsNonRoot: true runAsUser: 1000 capabilities: drop: - ALL ` ## Autoscaling `bash helm install aurora ./helm \ -n aurora --create-namespace \ --set autoscaling.enabled=true \ --set autoscaling.minReplicas=2 \ --set autoscaling.maxReplicas=10 \ --set autoscaling.targetCPUUtilizationPercentage=70 ` ## Verify `bash curl http://your-aurora-host/v1/models \ -H "Authorization: Bearer your-master-key" `` See the Helm chart README for the complete configuration reference.