clean up

Author: mitchross

docs/argocd.md

@@ -34,7 +34,7 @@ To solve the "chicken-and-egg" problem of bootstrapping a cluster (e.g., needing
 | Wave | Phase | Components | Description |
 |------|-------|------------|-------------|
 | **0** | **Foundation** | `cilium`, `1password-connect`, `external-secrets` | **Networking & Secrets**. The absolute minimum required for other pods to start and pull credentials. |
-| **1** | **Storage** | `longhorn`, `garage` | **Persistence**. Depends on Wave 0 for Pod-to-Pod communication and S3 backup credentials. |
+| **1** | **Storage** | `longhorn`, `snapshot-controller`, `volsync` | **Persistence**. Depends on Wave 0 for Pod-to-Pod communication and secrets. |
 | **2** | **System** | `cert-manager`, `gpu-operator`, `databases` | **Core Services**. Depends on Storage (PVCs) and Networking (Ingress/Gateway). |
 | **3** | **Observability** | `kube-prometheus-stack`, `loki` | **Monitoring**. Monitors the healthy stack. |
 | **4** | **User** | `my-apps/*` | **Workloads**. The actual applications running on the cluster. |

infrastructure/controllers/argocd/apps/garage.yaml

@@ -14,4 +14,3 @@ resources:
   - infrastructure-appset.yaml  # Wave 2
   - monitoring-appset.yaml      # Wave 3
   - my-apps-appset.yaml         # Wave 4
-  - garage.yaml

infrastructure/controllers/argocd/apps/kustomization.yaml

@@ -12,7 +12,6 @@ spec:
   description: "Infrastructure components like controllers, networking, and storage."
   sourceRepos:
   - https://github.com/mitchross/talos-argocd-proxmox.git
-  - https://git.deuxfleurs.fr/Deuxfleurs/garage.git
   destinations:
   - namespace: '*'
     server: https://kubernetes.default.svc

infrastructure/controllers/argocd/apps/projects.yaml

@@ -1,53 +1,3 @@
 # infrastructure/controllers/argocd/apps/values.yaml
 deployment:
   replicaCount: 1
-
-garage:
-  replicationFactor: "1"
-  rpcBindAddr: "[::]:3901"
-  # Generate these with: openssl rand -hex 32
-  admin_token: "2ec9bf4ca836db5d7ad57903d91bdb5e5365fc062fa74b228fc1fc916fef6ebe"
-  metrics_token: "2ec9bf4ca836db5d7ad57903d91bdb5e5365fc062fa74b228fc1fc916fef6ebe"
-  dbEngine: "lmdb"  # or "sqlite" if preferred
-
-  s3:
-    api:
-      region: "garage"
-      rootDomain: ".s3.garage.local"
-    web:
-      rootDomain: ".web.garage.local"
-      index: "index.html"
-
-persistence:
-  enabled: true
-  meta:
-    size: "1Gi"
-    storageClass: "longhorn"
-  data:
-    size: "10Gi"
-    storageClass: "longhorn"
-
-service:
-  type: ClusterIP
-  # Ensure admin API is exposed through the Service so in-cluster clients (webui)
-  # can reach port 3903. The upstream chart will render these ports if it uses
-  # a `service.ports` values structure. If the chart uses a different shape,
-  # adapt accordingly.
-  ports:
-    - name: s3-api
-      port: 3900
-      targetPort: 3900
-      protocol: TCP
-    - name: s3-web
-      port: 3902
-      targetPort: 3902
-      protocol: TCP
-    - name: admin
-      port: 3903
-      targetPort: 3903
-      protocol: TCP
-
-# Optional: enable metrics for Prometheus
-# monitoring:
-#   metrics:
-#     enabled: true

infrastructure/controllers/argocd/apps/values.yaml

@@ -167,7 +167,6 @@ machine:
 **HTTPRoutes:**
 - ArgoCD: `argocd.vanillax.me`
 - Longhorn: `longhorn.vanillax.me`
-- Garage (new): `garage.vanillax.me`
 - Many others...
 
 ## Network Flow Examples

infrastructure/networking/README.md

@@ -6,7 +6,7 @@ defaultSettings:
   # NOTE: Longhorn 1.10 consolidates settings for V1/V2 data engines. Use either
   # a single scalar or a JSON string with per-engine values like {"v1":"...","v2":"..."} when needed.
   defaultDataPath: "/var/lib/longhorn"
-  storageMinimalAvailablePercentage: "25"
+  storageMinimalAvailablePercentage: "10"
   storageOverProvisioningPercentage: "100"
   allowRecurringJobWhileVolumeDetached: "true"
   replicaAutoBalance: "best-effort"

infrastructure/storage/longhorn/values.yaml

@@ -0,0 +1,22 @@
+apiVersion: external-secrets.io/v1
+kind: ExternalSecret
+metadata:
+  name: loki-s3-credentials
+  namespace: loki-stack
+spec:
+  refreshInterval: "1h"
+  secretStoreRef:
+    kind: ClusterSecretStore
+    name: 1password
+  target:
+    name: loki-s3-credentials
+    creationPolicy: Owner
+  data:
+  - secretKey: AWS_ACCESS_KEY_ID
+    remoteRef:
+      key: rustfs
+      property: loki_access_key
+  - secretKey: AWS_SECRET_ACCESS_KEY
+    remoteRef:
+      key: rustfs
+      property: loki

monitoring/loki-stack/externalsecret.yaml

@@ -4,6 +4,7 @@ namespace: loki-stack
 resources:
   - ns.yaml
   - loki-http-route.yaml
+  - externalsecret.yaml
 helmCharts:
   - name: loki
     repo: https://grafana.github.io/helm-charts

monitoring/loki-stack/kustomization.yaml

@@ -1,15 +1,20 @@
 loki:
   commonConfig:
     path_prefix: /var/loki
-    replication_factor: 3
+    replication_factor: 2
   storage:
     type: "s3"
     bucketNames:
-      chunks: loki-chunks
-      ruler: loki-ruler
-      admin: loki-admin
+      chunks: loki
+      ruler: loki
+      admin: loki
     s3:
+      endpoint: http://192.168.10.133:30293
       region: us-east-1
+      secretAccessKey: ${AWS_SECRET_ACCESS_KEY}
+      accessKeyId: ${AWS_ACCESS_KEY_ID}
+      s3ForcePathStyle: true
+      insecure: true
   schemaConfig:
     configs:
     - from: "2024-04-01"
@@ -22,9 +27,7 @@ loki:
   ingester:
     chunk_encoding: snappy
   querier:
-    # Default is 4, if you have enough memory and CPU you can increase, reduce if OOMing
     max_concurrent: 4
-  # Configure query scheduler for read pods
   query_scheduler:
     use_scheduler_ring: true
   frontend_worker:
@@ -45,6 +48,9 @@ backend:
     limits:
       cpu: 500m
       memory: 1Gi
+  extraEnvFrom:
+  - secretRef:
+      name: loki-s3-credentials
 read:
   replicas: 2
   resources:
@@ -54,6 +60,9 @@ read:
     limits:
       cpu: 1000m
       memory: 1Gi
+  extraEnvFrom:
+  - secretRef:
+      name: loki-s3-credentials
 write:
   replicas: 3
   resources:
@@ -63,20 +72,13 @@ write:
     limits:
       cpu: 500m
       memory: 1Gi
-# Enable minio for storage
+  extraEnvFrom:
+  - secretRef:
+      name: loki-s3-credentials
+# DISABLED - Using external S3 (TrueNAS RustFS) instead
+# This saves ~200Gi of Longhorn storage
 minio:
-  enabled: true
-  persistence:
-    enabled: true
-    storageClass: longhorn
-    size: 50Gi
-  resources:
-    requests:
-      cpu: 100m
-      memory: 256Mi
-    limits:
-      cpu: 500m
-      memory: 1Gi
+  enabled: false
 gateway:
   service:
     type: LoadBalancer