feat(helm): rewrite values.yaml and update Chart.yaml for v1.0.0

Complete values.yaml with all config sections: API, frontend, poller, postgres, redis, NATS, OpenBao, WireGuard, storage, ingress, SMTP, secrets, telemetry, backup, and storage paths. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-17 18:40:57 -05:00
parent c8dbada606
commit fdf1735531
2 changed files with 239 additions and 56 deletions
--- a/infrastructure/helm/Chart.yaml
+++ b/infrastructure/helm/Chart.yaml
@@ -1,13 +1,17 @@
 apiVersion: v2
 name: tod
-description: The Other Dude — MikroTik Fleet Management
+description: The Other Dude — MikroTik fleet management platform
 type: application
-version: 9.0.1
+version: 1.0.0
 appVersion: "9.0.1"
+kubeVersion: ">=1.28.0-0"
 keywords:
  - mikrotik
-  - network-management
+  - network
  - fleet-management
+  - monitoring
 home: https://theotherdude.net
+sources:
+  - https://github.com/staack/the-other-dude
 maintainers:
-  - name: The Other Dude Team
+  - name: Jason Staack
--- a/infrastructure/helm/values.yaml
+++ b/infrastructure/helm/values.yaml
@@ -1,16 +1,20 @@
-# Default values for the-other-dude.
-# These values should work with `helm install` out of the box for development.
-# Production deployments MUST override secrets.jwtSecretKey, secrets.credentialEncryptionKey,
-# and secrets.firstAdminPassword.
+# Default values for TOD (The Other Dude).
+# Production deployments MUST override all empty secrets.* values.
+# See: https://theotherdude.net/docs/deployment/kubernetes
+
+# -- Global overrides
+nameOverride: ""
+fullnameOverride: ""

 # -----------------------------------------------------------------------
-# API service
+# API service (FastAPI + Gunicorn)
+# Serves the REST API, runs Alembic migrations on startup via init container.
 # -----------------------------------------------------------------------
 api:
  replicaCount: 1

  image:
-    repository: the-other-dude/api
+    repository: tod/api
    tag: latest
    pullPolicy: IfNotPresent

@@ -26,35 +30,52 @@ api:
      cpu: 500m
      memory: 512Mi

-  # Liveness and readiness probe configuration
  probes:
    liveness:
-      path: /api/health
+      path: /health
      initialDelaySeconds: 15
      periodSeconds: 10
      failureThreshold: 3
    readiness:
-      path: /api/health
+      path: /health/ready
      initialDelaySeconds: 10
      periodSeconds: 5
      failureThreshold: 3

  env:
-    # Token expiry (minutes for access, days for refresh)
-    jwtAccessTokenExpireMinutes: 15
-    jwtRefreshTokenExpireDays: 7
-    # CORS — set to your frontend origin in production
-    corsOrigins: "http://localhost:3000,http://localhost:5173"
+    # Runtime environment — set to "dev" for local testing (bypasses insecure-default checks)
+    environment: "production"
+    logLevel: "info"
    debug: "false"
+    # Number of Gunicorn worker processes
+    gunicornWorkers: "2"
+    # Comma-separated allowed CORS origins
+    corsOrigins: "http://localhost:3000"
+    # Public base URL for the app (used in emails, links). Leave empty for auto-detect.
+    appBaseUrl: ""
+    # JWT configuration
+    jwtAlgorithm: "HS256"
+    jwtAccessTokenExpireMinutes: "15"
+    jwtRefreshTokenExpireDays: "7"
+    # Database connection pool (leave empty for Python defaults)
+    dbPoolSize: ""
+    dbMaxOverflow: ""
+    dbAdminPoolSize: ""
+    dbAdminMaxOverflow: ""
+    # How often to check for new MikroTik firmware (hours)
+    firmwareCheckIntervalHours: ""
+    # Password reset token lifetime (minutes)
+    passwordResetTokenExpireMinutes: ""

 # -----------------------------------------------------------------------
-# Frontend service
+# Frontend service (React SPA served by nginx)
+# API URL is baked in at build time — no runtime env needed.
 # -----------------------------------------------------------------------
 frontend:
  replicaCount: 1

  image:
-    repository: the-other-dude/frontend
+    repository: tod/frontend
    tag: latest
    pullPolicy: IfNotPresent

@@ -70,21 +91,73 @@ frontend:
      cpu: 200m
      memory: 128Mi

+# -----------------------------------------------------------------------
+# Poller service (Go binary)
+# Polls MikroTik devices, manages SSH tunnels, reports to NATS.
+# -----------------------------------------------------------------------
+poller:
+  replicaCount: 1
+
+  image:
+    repository: tod/poller
+    tag: latest
+    pullPolicy: IfNotPresent
+
+  resources:
+    requests:
+      cpu: 50m
+      memory: 64Mi
+    limits:
+      cpu: 200m
+      memory: 256Mi
+
+  env:
+    # How often to poll each device (seconds)
+    pollIntervalSeconds: "60"
+    # SSH connection timeout to devices (seconds)
+    connectionTimeoutSeconds: "10"
+    # SSH command execution timeout (seconds)
+    commandTimeoutSeconds: "30"
+    # How often to refresh the device list from DB (seconds)
+    deviceRefreshSeconds: "60"
+    logLevel: "info"
+    # WinBox tunnel port range
+    tunnelPortMin: "49000"
+    tunnelPortMax: "49100"
+    # Idle timeout before closing a tunnel (seconds)
+    tunnelIdleTimeout: "300"
+    # SSH relay listener port
+    sshRelayPort: "8080"
+    # SSH session idle timeout (seconds)
+    sshIdleTimeout: "900"
+    # Max concurrent SSH sessions (global)
+    sshMaxSessions: "200"
+    # Max concurrent SSH sessions per user
+    sshMaxPerUser: "10"
+    # Max concurrent SSH sessions per device
+    sshMaxPerDevice: "20"
+    # Circuit breaker settings (leave empty for Go defaults)
+    circuitBreakerMaxFailures: ""
+    circuitBreakerBaseBackoffSeconds: ""
+    circuitBreakerMaxBackoffSeconds: ""
+
 # -----------------------------------------------------------------------
 # PostgreSQL (TimescaleDB)
+# Set enabled=false and provide externalUrl to use an existing database.
 # -----------------------------------------------------------------------
 postgres:
-  # Set to false to use an external PostgreSQL instance (provide externalUrl below)
  enabled: true

  image:
    repository: timescale/timescaledb
-    tag: latest-pg17
+    # Pinned for reproducibility — TimescaleDB 2.17.2 on PostgreSQL 17
+    tag: "2.17.2-pg17"
    pullPolicy: IfNotPresent

-  # Storage for the PVC
+  # PVC size for database files
  storage: 10Gi
-  storageClass: ""  # leave empty to use cluster default StorageClass
+  # Leave empty to use cluster default StorageClass
+  storageClass: ""

  service:
    port: 5432
@@ -92,23 +165,23 @@ postgres:
  auth:
    database: tod
    username: postgres
-    # password is sourced from secrets.dbPassword
+    # app_user: RLS-enforced role for tenant-scoped API queries
    appUsername: app_user
-    # appPassword is sourced from secrets.dbAppPassword

  resources:
    requests:
      cpu: 250m
      memory: 512Mi
    limits:
-      cpu: 1000m
+      cpu: "1"
      memory: 2Gi

  # External PostgreSQL URL (used when postgres.enabled=false)
  # externalUrl: "postgresql+asyncpg://user:pass@host:5432/tod"

 # -----------------------------------------------------------------------
-# Redis
+# Redis (ephemeral cache — no PVC)
+# Set enabled=false and provide externalUrl to use an existing instance.
 # -----------------------------------------------------------------------
 redis:
  enabled: true
@@ -127,10 +200,11 @@ redis:
      memory: 64Mi
    limits:
      cpu: 200m
-      memory: 256Mi
+      memory: 128Mi

 # -----------------------------------------------------------------------
-# NATS JetStream
+# NATS JetStream (message bus)
+# Set enabled=false and provide externalUrl to use an existing instance.
 # -----------------------------------------------------------------------
 nats:
  enabled: true
@@ -140,8 +214,9 @@ nats:
    tag: 2-alpine
    pullPolicy: IfNotPresent

+  # PVC size for JetStream storage
  storage: 5Gi
-  storageClass: ""  # leave empty to use cluster default StorageClass
+  storageClass: ""

  service:
    port: 4222
@@ -155,19 +230,64 @@ nats:
      memory: 512Mi

 # -----------------------------------------------------------------------
-# Go Poller
+# OpenBao (secrets engine)
+# Stores device credentials encrypted at rest. Must be manually initialized
+# and unsealed after first deploy and after every pod restart.
+# Set enabled=false and provide externalAddr to use an existing instance.
 # -----------------------------------------------------------------------
-poller:
-  replicaCount: 2
+openbao:
+  enabled: true

  image:
-    repository: the-other-dude/poller
+    repository: openbao/openbao
+    # Pinned — OpenBao 2.1
+    tag: "2.1"
+    pullPolicy: IfNotPresent
+
+  # PVC size for vault storage
+  storage: 1Gi
+  storageClass: ""
+
+  resources:
+    requests:
+      cpu: 50m
+      memory: 128Mi
+    limits:
+      cpu: 200m
+      memory: 256Mi
+
+# -----------------------------------------------------------------------
+# WireGuard VPN
+# Provides VPN tunnels for reaching devices behind NAT.
+# Set enabled=false if not using VPN features.
+# -----------------------------------------------------------------------
+wireguard:
+  enabled: true
+
+  image:
+    repository: lscr.io/linuxserver/wireguard
    tag: latest
    pullPolicy: IfNotPresent

+  # PVC size for WireGuard peer configs (shared with API)
+  storage: 1Gi
+  storageClass: ""
+
+  service:
+    # LoadBalancer exposes UDP directly; use NodePort if no cloud LB available
+    type: LoadBalancer
+    port: 51820
+
  env:
-    pollIntervalSeconds: 60
-    logLevel: info
+    # Public hostname or IP for WireGuard endpoint (required for VPN)
+    serverUrl: ""
+    serverPort: "51820"
+    # VPN device subnet
+    internalSubnet: "10.10.0.0/16"
+    # LinuxServer.io container UID/GID
+    puid: "1000"
+    pgid: "1000"
+    tz: "UTC"

  resources:
    requests:
@@ -175,45 +295,104 @@ poller:
      memory: 64Mi
    limits:
      cpu: 200m
-      memory: 256Mi
+      memory: 128Mi
+
+# -----------------------------------------------------------------------
+# Storage (standalone PVCs shared across services)
+# -----------------------------------------------------------------------
+storage:
+  gitStore:
+    # Config backup git repositories
+    size: 5Gi
+    storageClass: ""
+  firmwareCache:
+    # Cached MikroTik firmware downloads
+    size: 2Gi
+    storageClass: ""

 # -----------------------------------------------------------------------
 # Ingress
+# Routes HTTP traffic to frontend and API services.
+# WireGuard is exposed separately via its own LoadBalancer/NodePort service.
 # -----------------------------------------------------------------------
 ingress:
  enabled: true
  className: nginx
-  # annotations:
-  #   cert-manager.io/cluster-issuer: letsencrypt-prod
-
-  # host: tod.example.com  — set this in your deployment
+  # Required for production — set to your domain
  host: ""
-
  tls:
    enabled: false
-    # secretName: the-other-dude-tls
+    secretName: ""
+  # Additional annotations (e.g. cert-manager.io/cluster-issuer: letsencrypt-prod)
+  annotations: {}
+
+# -----------------------------------------------------------------------
+# SMTP (outbound email for password resets, alerts)
+# -----------------------------------------------------------------------
+smtp:
+  host: ""
+  port: "587"
+  useTls: "true"
+  fromAddress: "noreply@example.com"

 # -----------------------------------------------------------------------
 # Secrets
-# IMPORTANT: All secrets below MUST be overridden in production.
+# ALL empty values MUST be overridden in production.
+# Generate keys with:
+#   JWT:        openssl rand -hex 32
+#   Encryption: python -c "import secrets,base64; print(base64.b64encode(secrets.token_bytes(32)).decode())"
+#   Passwords:  openssl rand -base64 24
 # -----------------------------------------------------------------------
 secrets:
-  # JWT signing key — generate with: openssl rand -hex 32
+  # JWT signing key
  jwtSecretKey: ""
-
  # AES-256-GCM credential encryption key (base64-encoded 32 bytes)
-  # Generate with: python -c "import secrets, base64; print(base64.b64encode(secrets.token_bytes(32)).decode())"
  credentialEncryptionKey: ""
-
-  # First admin account (created on first startup)
+  # OpenBao root token (set after manual initialization)
+  openbaoToken: ""
+  # OpenBao unseal key (set after manual initialization)
+  baoUnsealKey: ""
+  # First admin account (created on initial startup)
  firstAdminEmail: "admin@the-other-dude.local"
  firstAdminPassword: ""
-
  # PostgreSQL superuser password
-  dbPassword: "postgres"
+  dbPassword: ""
+  # app_user password (RLS-enforced)
+  dbAppPassword: ""
+  # poller_user password (bypasses RLS)
+  dbPollerPassword: ""
+  # SMTP credentials
+  smtpUser: ""
+  smtpPassword: ""

-  # app_user password (non-superuser, RLS-enforced)
-  dbAppPassword: "app_password"
+# -----------------------------------------------------------------------
+# Telemetry (anonymous usage metrics)
+# Disabled by default. When enabled, sends anonymous stats to the collector.
+# -----------------------------------------------------------------------
+telemetry:
+  enabled: false
+  collectorUrl: "https://telemetry.theotherdude.net"

-  # poller_user password (bypasses RLS — SELECT on devices only)
-  dbPollerPassword: "poller_password"
+# -----------------------------------------------------------------------
+# Backup settings
+# -----------------------------------------------------------------------
+backup:
+  # How often to run config backups (seconds). Default: 6 hours.
+  configBackupInterval: "21600"
+  # Max concurrent backup jobs
+  configBackupMaxConcurrent: "10"
+  # Per-device backup command timeout (leave empty for default)
+  configBackupCommandTimeout: ""
+  # Days to retain config backup history
+  configRetentionDays: "90"
+
+# -----------------------------------------------------------------------
+# Storage paths (mounted inside containers)
+# These must match the PVC mount points in deployment templates.
+# -----------------------------------------------------------------------
+storagePaths:
+  gitStorePath: "/data/git-store"
+  firmwareCacheDir: "/data/firmware-cache"
+  wireguardConfigPath: "/data/wireguard"
+  # Hostname used by API route-setup init container to resolve WireGuard pod IP
+  wireguardGateway: "wireguard"