This Terraform module deploys Valkey (Redis fork) on Kubernetes using Helm charts. It provides a complete solution for deploying Valkey with optional clustering, sentinel support, monitoring, and dedicated node scheduling.

• Standalone and Cluster Modes: Deploy Valkey as standalone or with replication
• Sentinel Support: High availability with Redis Sentinel
• Authentication: Optional password-based authentication
• Monitoring: Built-in Prometheus metrics and ServiceMonitor support
• Dedicated Nodes: Schedule pods on specific nodes with custom labels and tolerations
• Persistent Storage: Configurable persistent volumes
• Resource Management: CPU and memory limits/requests
• Pod Disruption Budgets: Configurable PDB for high availability
• Custom Configuration: Valkey configuration via common_configuration variable

module "valkey" {
  source = "github.com/alexmorbo/terraform-helm-valkey"

  name                = "valkey"
  create_namespace    = true
  cluster_enabled     = true
  slave_count         = 3
  sentinel_enabled    = true
  use_password        = true
  storage_class       = "fast-ssd"
  metrics_enabled     = true
  service_monitor_enabled = true
}

module "valkey" {
  source = "github.com/alexmorbo/terraform-helm-valkey"

  name                = "valkey"
  create_namespace    = true
  cluster_enabled     = true
  slave_count         = 3
  sentinel_enabled    = true
  use_password        = true
  storage_class       = "fast-ssd"

  # Dedicated node scheduling
  dedicated_nodes     = true
  dedicated_node_group = "infra"
  node_group_label    = "company.com/group"
  dedicated_label     = "company.com/dedicated"

  # Resource configuration
  valkey_resources = {
    limits = {
      cpu    = "1000m"
      memory = "1Gi"
    }
    requests = {
      cpu    = "200m"
      memory = "256Mi"
    }
  }

  sentinel_resources = {
    limits = {
      cpu    = "200m"
      memory = "256Mi"
    }
    requests = {
      cpu    = "50m"
      memory = "64Mi"
    }
  }

  # Monitoring
  metrics_enabled = true
  service_monitor_enabled = true
  metrics_resources = {
    limits = {
      cpu    = "100m"
      memory = "128Mi"
    }
    requests = {
      cpu    = "50m"
      memory = "64Mi"
    }
  }
}

When dedicated_nodes = true, the module will:

1. Node Selector: Schedule pods only on nodes with the specified group label
2. Tolerations: Add tolerations for dedicated node taints
3. Customizable Labels: Use configurable label keys via node_group_label and dedicated_label

# Label nodes for group selection
kubectl label nodes node-1 company.com/group=infra
kubectl label nodes node-2 company.com/group=infra

# Taint nodes for dedicated scheduling
kubectl taint nodes node-1 company.com/dedicated=infra:NoExecute
kubectl taint nodes node-2 company.com/dedicated=infra:NoExecute

The module supports Prometheus monitoring with:

• Metrics Collection: Built-in metrics exporter
• ServiceMonitor: Automatic Prometheus target discovery
• Alerts: Pre-configured PrometheusRule for common issues:
  • Valkey instance down
  • AOF size threshold exceeded

• Password Authentication: Optional password protection
• Secret Management: Passwords stored in Kubernetes secrets
• Network Policies: Compatible with Kubernetes network policies

This module is released under the MIT License.