multi-cloud no mcp

2026-04-22 13:31:09 +02:00
parent f97b613c12
commit cab0866e14
54 changed files with 1150 additions and 83 deletions
--- a/README.md
+++ b/README.md
@@ -1,9 +1,9 @@
 # Kubernetes Cluster - GitOps Configuration

-> **Kubernetes cluster bootstrapping and GitOps configuration repository** using ArgoCD for UpCloud Managed Kubernetes
+> **Kubernetes cluster bootstrapping and GitOps configuration repository** using ArgoCD for multi-cloud Kubernetes (UpCloud, AWS EKS, Azure AKS, GCP GKE)

 [![GitOps](https://img.shields.io/badge/GitOps-ArgoCD-blue)](https://argoproj.github.io/cd/)
-[![Kubernetes](https://img.shields.io/badge/Kubernetes-UpCloud-orange)](https://upcloud.com/)
+[![Kubernetes](https://img.shields.io/badge/Kubernetes-Multi--Cloud-orange)]()

 ---

@@ -95,14 +95,26 @@ This repository contains the complete GitOps configuration for our Kubernetes cl
 │   │   ├── renovate.yaml
 │   │   ├── ...                # All other Application manifests
 │   │   └── secrets.yaml
-│   ├── overlays/              # Per-cluster overrides
-│   │   ├── upc-dev/           # UpCloud Dev cluster (uses base as-is)
-│   │   └── upc-prod/         # UpCloud Prod cluster (patches value paths)
+│   ├── overlays/              # Per-cluster overrides (Kustomize)
+│   │   ├── upc-dev/           # UpCloud Dev (uses base as-is)
+│   │   ├── upc-prod/         # UpCloud Prod (patches value paths)
+│   │   ├── aws-dev/           # AWS EKS Dev
+│   │   ├── aws-prod/         # AWS EKS Prod
+│   │   ├── azure-dev/        # Azure AKS Dev
+│   │   ├── azure-prod/       # Azure AKS Prod
+│   │   ├── gcp-dev/          # GCP GKE Dev
+│   │   └── gcp-prod/         # GCP GKE Prod
 │   ├── dashboards/            # Grafana dashboard ConfigMaps
 │   └── values/                # Helm value overrides
-│       ├── base/              # Shared values (all clusters)
-│       ├── upc-dev/           # UpCloud Dev-specific values
-│       └── upc-prod/         # UpCloud Prod-specific values
+│       ├── base/              # Shared cloud-agnostic values
+│       ├── upc-dev/           # UpCloud Dev (storage, LB, pricing)
+│       ├── upc-prod/         # UpCloud Prod
+│       ├── aws-dev/           # AWS EKS Dev
+│       ├── aws-prod/         # AWS EKS Prod
+│       ├── azure-dev/        # Azure AKS Dev
+│       ├── azure-prod/       # Azure AKS Prod
+│       ├── gcp-dev/          # GCP GKE Dev
+│       └── gcp-prod/         # GCP GKE Prod
 │
 ├── apps/                      # Business Applications
 │   ├── mcp10x.yaml
@@ -361,7 +373,7 @@ kubectl patch application myapp -n argocd \
 ## 📖 Key Concepts

 ### App-of-Apps Pattern
-`_app-of-apps.yaml` is the root Application that manages all other Applications in `infra/`. Kustomize overlays in `infra/overlays/{upc-dev,upc-prod}/` render the base Applications with per-cluster patches (e.g., swapping value file paths from `upc-dev` to `upc-prod`).
+`_app-of-apps-{cluster}.yaml` is the root Application that manages all other Applications in `infra/`. Kustomize overlays in `infra/overlays/{cluster}/` render the base Applications with per-cluster patches (e.g., swapping value file paths). Supported clusters: `upc-dev`, `upc-prod`, `aws-dev`, `aws-prod`, `azure-dev`, `azure-prod`, `gcp-dev`, `gcp-prod`.

 ### Multi-Source Pattern
 Applications reference both:
@@ -458,16 +470,14 @@ Documentation lives in `docs/`. To update:
 ## 📝 Notes

 ### Current Environment
- **Provider**: UpCloud Managed Kubernetes
+- **Provider**: Multi-cloud (UpCloud, AWS EKS, Azure AKS, GCP GKE)
+- **Active clusters**: UpCloud (upc-dev, upc-prod)
 - **Environment**: Production (internal use only)
- **Clusters**: Multi-cluster (upc-dev, upc-prod) via Kustomize overlays
 - **Auth**: Disabled for ArgoCD (internal access)
- **Backup**: None (cluster rebuildable via GitOps)
+- **Backup**: Gitea daily backup to S3-compatible storage

 ### Known Limitations
- No automated backups (yet)
 - Secret rotation not automated
- Multi-cluster limited to upc-dev and upc-prod environments
 - DNS management is manual

 **Future improvements**: See [Operations Runbook - Disaster Recovery](docs/OPERATIONS-RUNBOOK.md#disaster-recovery)
@@ -504,7 +514,7 @@ Internal use only. Not for public distribution.

 ---

-**Last Updated**: 2026-03-16
+**Last Updated**: 2026-04-22
 **Documentation Version**: 1.0.0

 **🚀 Ready to get started? Check out the [Documentation Index](docs/README.md)!**
--- a/_app-of-apps-aws-dev.yaml
+++ b/_app-of-apps-aws-dev.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: monitoring
+  annotations:
+    argocd.argoproj.io/sync-wave: "-1"
+---
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: infrastructure-apps
+  namespace: argocd
+  labels:
+    app.kubernetes.io/name: infrastructure-apps
+    app.kubernetes.io/part-of: platform
+  finalizers:
+  - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  source:
+    repoURL: git@github.com:fortedigital/sturdy-adventure.git
+    targetRevision: HEAD
+    path: infra/overlays/aws-dev
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: default
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true
+    syncOptions:
+    - CreateNamespace=true
--- a/_app-of-apps-aws-prod.yaml
+++ b/_app-of-apps-aws-prod.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: monitoring
+  annotations:
+    argocd.argoproj.io/sync-wave: "-1"
+---
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: infrastructure-apps
+  namespace: argocd
+  labels:
+    app.kubernetes.io/name: infrastructure-apps
+    app.kubernetes.io/part-of: platform
+  finalizers:
+  - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  source:
+    repoURL: git@github.com:fortedigital/sturdy-adventure.git
+    targetRevision: HEAD
+    path: infra/overlays/aws-prod
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: default
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true
+    syncOptions:
+    - CreateNamespace=true
--- a/_app-of-apps-azure-dev.yaml
+++ b/_app-of-apps-azure-dev.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: monitoring
+  annotations:
+    argocd.argoproj.io/sync-wave: "-1"
+---
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: infrastructure-apps
+  namespace: argocd
+  labels:
+    app.kubernetes.io/name: infrastructure-apps
+    app.kubernetes.io/part-of: platform
+  finalizers:
+  - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  source:
+    repoURL: git@github.com:fortedigital/sturdy-adventure.git
+    targetRevision: HEAD
+    path: infra/overlays/azure-dev
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: default
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true
+    syncOptions:
+    - CreateNamespace=true
--- a/_app-of-apps-azure-prod.yaml
+++ b/_app-of-apps-azure-prod.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: monitoring
+  annotations:
+    argocd.argoproj.io/sync-wave: "-1"
+---
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: infrastructure-apps
+  namespace: argocd
+  labels:
+    app.kubernetes.io/name: infrastructure-apps
+    app.kubernetes.io/part-of: platform
+  finalizers:
+  - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  source:
+    repoURL: git@github.com:fortedigital/sturdy-adventure.git
+    targetRevision: HEAD
+    path: infra/overlays/azure-prod
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: default
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true
+    syncOptions:
+    - CreateNamespace=true
--- a/_app-of-apps-gcp-dev.yaml
+++ b/_app-of-apps-gcp-dev.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: monitoring
+  annotations:
+    argocd.argoproj.io/sync-wave: "-1"
+---
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: infrastructure-apps
+  namespace: argocd
+  labels:
+    app.kubernetes.io/name: infrastructure-apps
+    app.kubernetes.io/part-of: platform
+  finalizers:
+  - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  source:
+    repoURL: git@github.com:fortedigital/sturdy-adventure.git
+    targetRevision: HEAD
+    path: infra/overlays/gcp-dev
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: default
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true
+    syncOptions:
+    - CreateNamespace=true
--- a/_app-of-apps-gcp-prod.yaml
+++ b/_app-of-apps-gcp-prod.yaml
@@ -0,0 +1,32 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: monitoring
+  annotations:
+    argocd.argoproj.io/sync-wave: "-1"
+---
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: infrastructure-apps
+  namespace: argocd
+  labels:
+    app.kubernetes.io/name: infrastructure-apps
+    app.kubernetes.io/part-of: platform
+  finalizers:
+  - resources-finalizer.argocd.argoproj.io
+spec:
+  project: default
+  source:
+    repoURL: git@github.com:fortedigital/sturdy-adventure.git
+    targetRevision: HEAD
+    path: infra/overlays/gcp-prod
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: default
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true
+    syncOptions:
+    - CreateNamespace=true
--- a/cluster-resources/gitea-backup-cronjob.yaml
+++ b/cluster-resources/gitea-backup-cronjob.yaml
@@ -57,17 +57,17 @@ spec:
                - sh
                - -c
                - |
-                  mc alias set upcloud "${S3_ENDPOINT}" "${AWS_ACCESS_KEY_ID}" "${AWS_SECRET_ACCESS_KEY}"
+                  mc alias set s3 "${S3_ENDPOINT}" "${AWS_ACCESS_KEY_ID}" "${AWS_SECRET_ACCESS_KEY}"

                  TIMESTAMP=$(date +%Y%m%d-%H%M%S)
                  KEY="gitea-dump-${TIMESTAMP}.zip"
                  echo "Uploading ${KEY}..."
-                  mc cp /backup/gitea-dump.zip "upcloud/${S3_BUCKET}/${KEY}" && \
+                  mc cp /backup/gitea-dump.zip "s3/${S3_BUCKET}/${KEY}" && \
                  echo "Upload complete."

                  # Prune backups older than 7 days
                  echo "Pruning backups older than 7 days..."
-                  mc rm --older-than 7d --force "upcloud/${S3_BUCKET}/" 2>&1 || true
+                  mc rm --older-than 7d --force "s3/${S3_BUCKET}/" 2>&1 || true
                  echo "Pruning complete."
              envFrom:
                - secretRef:
--- a/clusters/aws-dev.yaml
+++ b/clusters/aws-dev.yaml
@@ -0,0 +1,10 @@
+clusterName: dev-eks                  # <- adjust to your EKS cluster name
+domain: example.com                   # <- adjust to your domain
+argocdDomain: argocd.example.com
+grafanaDomain: grafana.example.com
+keycloakDomain: id.example.com
+dotaiDomain: kubemcp.example.com
+dotaiUiDomain: kubemcpui.example.com
+letsencryptEmail: admin@example.com   # <- adjust
+trustedIPs: "10.0.0.0/8"             # <- adjust to your VPC CIDR
+cloudProvider: aws
--- a/clusters/aws-prod.yaml
+++ b/clusters/aws-prod.yaml
@@ -0,0 +1,10 @@
+clusterName: prod-eks                 # <- adjust to your EKS cluster name
+domain: example.com                   # <- adjust to your domain
+argocdDomain: argocd.example.com
+grafanaDomain: grafana.example.com
+keycloakDomain: id.example.com
+dotaiDomain: kubemcp.example.com
+dotaiUiDomain: kubemcpui.example.com
+letsencryptEmail: admin@example.com   # <- adjust
+trustedIPs: "10.0.0.0/8"             # <- adjust to your VPC CIDR
+cloudProvider: aws
--- a/clusters/azure-dev.yaml
+++ b/clusters/azure-dev.yaml
@@ -0,0 +1,10 @@
+clusterName: dev-aks                  # <- adjust to your AKS cluster name
+domain: example.com                   # <- adjust to your domain
+argocdDomain: argocd.example.com
+grafanaDomain: grafana.example.com
+keycloakDomain: id.example.com
+dotaiDomain: kubemcp.example.com
+dotaiUiDomain: kubemcpui.example.com
+letsencryptEmail: admin@example.com   # <- adjust
+trustedIPs: "10.0.0.0/8,168.63.129.16/32"  # <- VNet CIDR + Azure health probe
+cloudProvider: azure
--- a/clusters/azure-prod.yaml
+++ b/clusters/azure-prod.yaml
@@ -0,0 +1,10 @@
+clusterName: prod-aks                 # <- adjust to your AKS cluster name
+domain: example.com                   # <- adjust to your domain
+argocdDomain: argocd.example.com
+grafanaDomain: grafana.example.com
+keycloakDomain: id.example.com
+dotaiDomain: kubemcp.example.com
+dotaiUiDomain: kubemcpui.example.com
+letsencryptEmail: admin@example.com   # <- adjust
+trustedIPs: "10.0.0.0/8,168.63.129.16/32"  # <- VNet CIDR + Azure health probe
+cloudProvider: azure
--- a/clusters/gcp-dev.yaml
+++ b/clusters/gcp-dev.yaml
@@ -0,0 +1,10 @@
+clusterName: dev-gke                  # <- adjust to your GKE cluster name
+domain: example.com                   # <- adjust to your domain
+argocdDomain: argocd.example.com
+grafanaDomain: grafana.example.com
+keycloakDomain: id.example.com
+dotaiDomain: kubemcp.example.com
+dotaiUiDomain: kubemcpui.example.com
+letsencryptEmail: admin@example.com   # <- adjust
+trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"  # <- subnet CIDR + GCP health checks
+cloudProvider: gcp
--- a/clusters/gcp-prod.yaml
+++ b/clusters/gcp-prod.yaml
@@ -0,0 +1,10 @@
+clusterName: prod-gke                 # <- adjust to your GKE cluster name
+domain: example.com                   # <- adjust to your domain
+argocdDomain: argocd.example.com
+grafanaDomain: grafana.example.com
+keycloakDomain: id.example.com
+dotaiDomain: kubemcp.example.com
+dotaiUiDomain: kubemcpui.example.com
+letsencryptEmail: admin@example.com   # <- adjust
+trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"  # <- subnet CIDR + GCP health checks
+cloudProvider: gcp
--- a/docs/GITOPS-ARCHITECTURE.md
+++ b/docs/GITOPS-ARCHITECTURE.md
@@ -12,11 +12,11 @@

 ## Overview

-This Kubernetes cluster uses a **GitOps approach** powered by **ArgoCD**, where Git repositories serve as the single source of truth for both infrastructure and application deployments. The cluster is running on **UpCloud Managed Kubernetes** but is designed to be cloud-agnostic.
+This Kubernetes cluster uses a **GitOps approach** powered by **ArgoCD**, where Git repositories serve as the single source of truth for both infrastructure and application deployments. The cluster setup is **cloud-agnostic**, with ready-to-use configurations for **UpCloud**, **AWS EKS**, **Azure AKS**, and **GCP GKE**.

 ### Key Characteristics
 - **Environment**: Production (internal use only)
- **Cluster Type**: Multi-cluster (upc-dev, upc-prod) via Kustomize overlays
+- **Cluster Type**: Multi-cloud, multi-cluster via Kustomize overlays (UpCloud, AWS, Azure, GCP)
 - **GitOps Tool**: ArgoCD
 - **Deployment Pattern**: App-of-Apps
 - **Secret Management**: Sealed Secrets (kubeseal)
@@ -63,7 +63,7 @@ This Kubernetes cluster uses a **GitOps approach** powered by **ArgoCD**, where
                                                        ▼
                                           ┌────────────────────────────────┐
                                           │   Kubernetes Clusters          │
-                                           │   (UpCloud: upc-dev, upc-prod) │
+                                           │   (UpCloud, AWS, Azure, GCP)    │
                                           │                                │
                                           │  ┌──────────────────────────┐  │
                                           │  │    ArgoCD                │  │
@@ -131,26 +131,22 @@ launchpad/
 │   │   ├── renovate.yaml
 │   │   ├── ...                       # All other Application manifests
 │   │   └── secrets.yaml
-│   ├── overlays/                     # Per-cluster overrides
+│   ├── overlays/                     # Per-cluster Kustomize overrides
 │   │   ├── upc-dev/                  # UpCloud Dev (uses base as-is)
-│   │   └── upc-prod/                # UpCloud Prod (patches value paths)
+│   │   ├── upc-prod/                # UpCloud Prod (patches value paths)
+│   │   ├── aws-dev/                  # AWS EKS Dev
+│   │   ├── aws-prod/                # AWS EKS Prod
+│   │   ├── azure-dev/               # Azure AKS Dev
+│   │   ├── azure-prod/              # Azure AKS Prod
+│   │   ├── gcp-dev/                 # GCP GKE Dev
+│   │   └── gcp-prod/               # GCP GKE Prod
 │   ├── dashboards/                   # Grafana dashboard ConfigMaps
 │   └── values/                       # Helm value overrides for infra
-│       ├── base/                     # Shared values (all clusters)
-│       │   ├── traefik-values.yaml
-│       │   ├── keycloak-values.yaml
-│       │   ├── grafana-values.yaml
-│       │   ├── prometheus-values.yaml
-│       │   ├── gitea-values.yaml
-│       │   └── ...
-│       ├── upc-dev/                  # upc-dev cluster-specific values
-│       │   ├── traefik-values.yaml
-│       │   ├── keycloak-values.yaml
-│       │   └── grafana-values.yaml
-│       └── upc-prod/                # upc-prod cluster-specific values
-│           ├── traefik-values.yaml
-│           ├── keycloak-values.yaml
-│           └── grafana-values.yaml
+│       ├── base/                     # Cloud-agnostic shared values
+│       ├── upc-{dev,prod}/           # UpCloud: storage class, LB, pricing
+│       ├── aws-{dev,prod}/           # AWS: gp3, NLB, CUR pricing
+│       ├── azure-{dev,prod}/         # Azure: managed-csi-premium, Standard LB
+│       └── gcp-{dev,prod}/           # GCP: premium-rwo, L4 LB
 │
 ├── apps/                             # Business Application ArgoCD manifests (Kustomize)
 │   ├── base/                         # Base app manifests
@@ -287,7 +283,7 @@ app-repository/
 ### The App-of-Apps Pattern

 ```
-_app-of-apps-{upc-dev,upc-prod}.yaml (Root, per cluster)
+_app-of-apps-{cluster}.yaml (Root, per cluster — e.g. upc-dev, aws-prod, gcp-dev)
    │
    ├── infrastructure-apps (manages infra/)
    │   ├── cluster-resources-application
@@ -377,6 +373,15 @@ patches:
      value: $values/infra/values/upc-prod/traefik-values.yaml
 ```

+Cloud-specific values (storage classes, load balancer annotations, cost model) are isolated in per-cluster value files. Base values are fully cloud-agnostic:
+
+| Cloud | Storage Class | Load Balancer | OpenCost Provider |
+|-------|--------------|---------------|-------------------|
+| **UpCloud** | `upcloud-block-storage-maxiops` | UpCloud LB (ProxyProtocol v2) | Custom pricing |
+| **AWS EKS** | `gp3` (EBS CSI) | NLB (ProxyProtocol v2) | AWS CUR |
+| **Azure AKS** | `managed-csi-premium` | Standard LB (`externalTrafficPolicy: Local`) | Azure Billing API |
+| **GCP GKE** | `premium-rwo` (PD CSI) | L4 passthrough NLB | GCP Cloud Billing |
+
 **Benefits**:
 - Single source of truth for Application definitions
 - Cluster-specific values isolated per overlay
@@ -658,6 +663,6 @@ Notifications include:

 ---

-**Last Updated**: 2026-03-16
+**Last Updated**: 2026-04-22
 **Maintained By**: Platform Team
 **Questions?**: Contact #platform-support on Slack
--- a/docs/OPERATIONS-RUNBOOK.md
+++ b/docs/OPERATIONS-RUNBOOK.md
@@ -37,7 +37,7 @@ Bootstrap a new cluster from scratch:

 #### Prerequisites

-1. **Kubernetes cluster running** (UpCloud or any K8s cluster)
+1. **Kubernetes cluster running** (UpCloud, AWS EKS, Azure AKS, GCP GKE, or any K8s cluster)
 2. **kubectl configured** with admin access
 3. **Repositories cloned** locally

@@ -54,11 +54,13 @@ kubectl get nodes
 git clone https://git.forteapps.net/Forte/launchpad
 cd launchpad

-# 2. Set cluster name (optional)
-export CLUSTER_NAME="prod-cluster-01"
+# 2. Run bootstrap script with cluster target
+# Available clusters: upc-dev, upc-prod, aws-dev, aws-prod,
+#                     azure-dev, azure-prod, gcp-dev, gcp-prod
+./bootstrap.sh upc-dev

-# 3. Run bootstrap script
-./bootstrap.sh
+# Cluster config is loaded from clusters/<cluster>.yaml
+# (cloudProvider, trustedIPs, domain, etc.)
 ```

 **What Happens:**
@@ -1262,13 +1264,21 @@ spec:

 ### Backup Strategy

-**Current State**: No automated backups
+**Current State**: Gitea daily backups to S3-compatible storage

-**What Needs Backup**:
- ❌ Cluster state (not backed up - recreate via GitOps)
- ❌ Persistent volumes (currently not critical)
- ✅ Git repositories (Gitea provides backup)
- ⚠️ Secrets (sealed secrets in Git, unseal keys need safekeeping)
+**What Is Backed Up**:
+- ✅ Gitea repositories + database: Daily CronJob (`cluster-resources/gitea-backup-cronjob.yaml`) uploads to S3-compatible storage with 7-day retention
+- ✅ Git repositories: Full cluster config recoverable from Git
+- ⚠️ Secrets: Sealed secrets in Git; unseal keys need safekeeping
+
+**What Is NOT Backed Up**:
+- ❌ Cluster state (recreate via GitOps)
+- ❌ Other persistent volumes (Prometheus, Loki, Tempo data)
+
+**Per-cloud backup scripts** (manual restore helpers):
+- UpCloud/AWS: `scripts/gitea-backup.sh` / `scripts/gitea-backup-aws.sh` (MinIO CLI, S3-compatible)
+- Azure: `scripts/gitea-backup-azure.sh` (Azure CLI + Blob Storage)
+- GCP: `scripts/gitea-backup-gcp.sh` (gsutil + GCS)

 ### Cluster Rebuild

@@ -1370,6 +1380,9 @@ kubectl get pods -n argocd

 ```bash
 # UpCloud: Upgrade via control panel or CLI
+# AWS EKS: eksctl upgrade cluster / AWS Console
+# Azure AKS: az aks upgrade / Azure Portal
+# GCP GKE: gcloud container clusters upgrade / Cloud Console

 # After upgrade, verify cluster
 kubectl version
@@ -1507,18 +1520,35 @@ git push

 ### Multi-Cluster Setup

-The repository supports multiple clusters via Kustomize overlays:
+The repository supports multiple clusters across multiple clouds via Kustomize overlays:

+**Active clusters:**
 - **upc-dev** (default): `infra/overlays/upc-dev/` — uses base Applications as-is
 - **upc-prod**: `infra/overlays/upc-prod/` — patches value file paths from `upc-dev` to `upc-prod`

-Each cluster has its own:
- Root app-of-apps file: `_app-of-apps-upc-dev.yaml` / `_app-of-apps-upc-prod.yaml`
- Cluster-specific Helm values: `infra/values/upc-dev/` / `infra/values/upc-prod/`
- Sealed secrets: `secrets/upc-dev/` (others as needed)
- Apps overlay: `apps/overlays/upc-dev/` / `apps/overlays/upc-prod/`
+**Cloud-ready templates (fill in `clusters/*.yaml` before use):**
+- **aws-dev** / **aws-prod**: AWS EKS with NLB, gp3 storage, AWS CUR pricing
+- **azure-dev** / **azure-prod**: Azure AKS with Standard LB, managed-csi-premium storage
+- **gcp-dev** / **gcp-prod**: GCP GKE with L4 LB, premium-rwo storage

-To add a new cluster, create a new overlay directory (e.g., `infra/overlays/upc-staging/`) with patches that swap the value file paths.
+Each cluster has its own:
+- Root app-of-apps: `_app-of-apps-{cluster}.yaml`
+- Cluster config: `clusters/{cluster}.yaml` (domain, trustedIPs, cloudProvider)
+- Kustomize overlay: `infra/overlays/{cluster}/kustomization.yaml`
+- Helm value overrides: `infra/values/{cluster}/` (traefik, gitea, opencost)
+- Sealed secrets: `secrets/{cluster}/` (as needed)
+- Apps overlay: `apps/overlays/{cluster}/`
+
+Cloud-specific values handled per-cluster:
+
+| Concern | UpCloud | AWS EKS | Azure AKS | GCP GKE |
+|---------|---------|---------|-----------|---------|
+| **Storage class** | `upcloud-block-storage-maxiops` | `gp3` | `managed-csi-premium` | `premium-rwo` |
+| **Load balancer** | UpCloud LB + ProxyProtocol v2 | NLB + ProxyProtocol v2 | Standard LB + `externalTrafficPolicy: Local` | L4 passthrough NLB |
+| **Cost monitoring** | Custom pricing | AWS CUR | Azure Billing API | GCP Cloud Billing |
+| **Backup storage** | UpCloud S3-compat | AWS S3 (native) | Azure Blob Storage | GCS |
+
+To add a new cluster, create a new overlay directory (e.g., `infra/overlays/aws-staging/`) with patches that swap the value file paths, and a matching `clusters/aws-staging.yaml`.

 ### Blue-Green Deployments

@@ -1661,6 +1691,6 @@ echo "Remember to delete: $SECRET_FILE"

 ---

-**Last Updated**: 2026-03-16
+**Last Updated**: 2026-04-22
 **Maintained By**: Platform Team
 **Emergency Contact**: #platform-support on Slack
--- a/docs/README.md
+++ b/docs/README.md
@@ -180,7 +180,7 @@ Reference for:
                            │
                            ▼
 ┌──────────────────────────────────────────────────────────────┐
-│          Kubernetes Clusters (UpCloud: upc-dev, upc-prod)     │
+│          Kubernetes Clusters (UpCloud, AWS, Azure, GCP)       │
 │  ┌──────────────────────────────────────────────────────┐    │
 │  │  Infrastructure: Traefik, Cert-Manager, Kyverno     │    │
 │  ├──────────────────────────────────────────────────────┤    │
@@ -194,7 +194,7 @@ Reference for:
 ### Key Technologies

 - **GitOps**: ArgoCD
- **Kubernetes**: UpCloud Managed Kubernetes (multi-cluster: upc-dev, upc-prod)
+- **Kubernetes**: Multi-cloud (UpCloud, AWS EKS, Azure AKS, GCP GKE)
 - **Ingress**: Traefik v2
 - **Certificates**: Cert-Manager + Let's Encrypt
 - **Policies**: Kyverno
@@ -299,11 +299,16 @@ docs/
 ## 🔄 Documentation Versions

 **Current Version**: 1.0.0
-**Last Updated**: 2026-03-16
+**Last Updated**: 2026-04-22
 **Maintained By**: Platform Team

 ### Changelog

+- **v1.1.0 (2026-04-22)**: Multi-cloud support
+  - Cloud-agnostic base values (storage, LB, pricing moved to per-cluster overlays)
+  - Added AWS EKS, Azure AKS, GCP GKE configurations
+  - Per-cloud backup scripts
+  - Updated all documentation
 - **v1.0.0 (2026-03-16)**: Initial comprehensive documentation release
  - GitOps Architecture guide
  - Developer Onboarding guide
--- a/docs/REFERENCE.md
+++ b/docs/REFERENCE.md
@@ -19,9 +19,9 @@

 | Component | Value |
 |-----------|-------|
-| **Provider** | UpCloud Managed Kubernetes |
-| **Environment** | Production (internal use) |
-| **Cluster Count** | Multi-cluster (upc-dev, upc-prod) |
+| **Provider** | Multi-cloud (UpCloud, AWS EKS, Azure AKS, GCP GKE) |
+| **Active clusters** | UpCloud (upc-dev, upc-prod) |
+| **Cloud-ready templates** | AWS, Azure, GCP (dev + prod each) |
 | **GitOps Tool** | ArgoCD |
 | **Ingress Controller** | Traefik v2 |
 | **Certificate Management** | Cert-Manager + Let's Encrypt |
@@ -42,7 +42,7 @@ Internet
 [DNS: *.forteapps.net]
   │
   ▼
-[UpCloud LoadBalancer]
+[Cloud Load Balancer]
   │
   ▼
 [Traefik Ingress Controller]
@@ -1470,14 +1470,22 @@ Recommended resource allocation:

 ### Storage Classes

-Default storage class used: **UpCloud default** (varies by provider)
+Storage classes are cloud-specific and configured in per-cluster value overrides (`infra/values/{cluster}/gitea-values.yaml`):
+
+| Cloud | Storage Class | Driver |
+|-------|--------------|--------|
+| **UpCloud** | `upcloud-block-storage-maxiops` | UpCloud CSI |
+| **AWS EKS** | `gp3` | EBS CSI |
+| **Azure AKS** | `managed-csi-premium` | Azure Disk CSI |
+| **GCP GKE** | `premium-rwo` | PD CSI |

 ```yaml
+# Example: base values omit storageClass (set in per-cluster overlay)
 persistence:
  enabled: true
-  storageClass: ""     # Uses default
  accessMode: ReadWriteOnce
  size: 5Gi
+  # storageClass set by infra/values/{cluster}/gitea-values.yaml
 ```

 ---
@@ -1673,6 +1681,6 @@ team: platform

 ---

-**Last Updated**: 2026-04-16
+**Last Updated**: 2026-04-22
 **Maintained By**: Platform Team
 **Version**: 1.0.0
--- a/infra/base/gitea.yaml
+++ b/infra/base/gitea.yaml
@@ -22,6 +22,7 @@ spec:
      releaseName: gitea
      valueFiles:
      - $values/infra/values/base/gitea-values.yaml
+      - $values/infra/values/upc-dev/gitea-values.yaml

  - repoURL: ssh://git@git.forteapps.net:2222/Forte/launchpad.git
    targetRevision: HEAD
--- a/infra/base/opencost.yaml
+++ b/infra/base/opencost.yaml
@@ -22,6 +22,7 @@ spec:
      releaseName: opencost
      valueFiles:
      - $values/infra/values/base/opencost-values.yaml
+      - $values/infra/values/upc-dev/opencost-values.yaml

  - repoURL: git@github.com:fortedigital/sturdy-adventure.git
    targetRevision: HEAD
--- a/infra/overlays/aws-dev/kustomization.yaml
+++ b/infra/overlays/aws-dev/kustomization.yaml
@@ -0,0 +1,35 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- ../../base
+
+patches:
+# Traefik: swap upc-dev → aws-dev
+- target:
+    kind: Application
+    name: traefik
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/aws-dev/traefik-values.yaml
+
+# Gitea: swap upc-dev → aws-dev
+- target:
+    kind: Application
+    name: gitea
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/aws-dev/gitea-values.yaml
+
+# OpenCost: swap upc-dev → aws-dev
+- target:
+    kind: Application
+    name: opencost
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/aws-dev/opencost-values.yaml
+
+# TODO: add patches for keycloak, grafana, secrets, enterprise-apps
+# when deploying to this cluster (these are deployment-specific, not cloud-specific)
--- a/infra/overlays/aws-prod/kustomization.yaml
+++ b/infra/overlays/aws-prod/kustomization.yaml
@@ -0,0 +1,35 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- ../../base
+
+patches:
+# Traefik: swap upc-dev → aws-prod
+- target:
+    kind: Application
+    name: traefik
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/aws-prod/traefik-values.yaml
+
+# Gitea: swap upc-dev → aws-prod
+- target:
+    kind: Application
+    name: gitea
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/aws-prod/gitea-values.yaml
+
+# OpenCost: swap upc-dev → aws-prod
+- target:
+    kind: Application
+    name: opencost
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/aws-prod/opencost-values.yaml
+
+# TODO: add patches for keycloak, grafana, secrets, enterprise-apps
+# when deploying to this cluster (these are deployment-specific, not cloud-specific)
--- a/infra/overlays/azure-dev/kustomization.yaml
+++ b/infra/overlays/azure-dev/kustomization.yaml
@@ -0,0 +1,35 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- ../../base
+
+patches:
+# Traefik: swap upc-dev → azure-dev
+- target:
+    kind: Application
+    name: traefik
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/azure-dev/traefik-values.yaml
+
+# Gitea: swap upc-dev → azure-dev
+- target:
+    kind: Application
+    name: gitea
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/azure-dev/gitea-values.yaml
+
+# OpenCost: swap upc-dev → azure-dev
+- target:
+    kind: Application
+    name: opencost
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/azure-dev/opencost-values.yaml
+
+# TODO: add patches for keycloak, grafana, secrets, enterprise-apps
+# when deploying to this cluster (these are deployment-specific, not cloud-specific)
--- a/infra/overlays/azure-prod/kustomization.yaml
+++ b/infra/overlays/azure-prod/kustomization.yaml
@@ -0,0 +1,35 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- ../../base
+
+patches:
+# Traefik: swap upc-dev → azure-prod
+- target:
+    kind: Application
+    name: traefik
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/azure-prod/traefik-values.yaml
+
+# Gitea: swap upc-dev → azure-prod
+- target:
+    kind: Application
+    name: gitea
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/azure-prod/gitea-values.yaml
+
+# OpenCost: swap upc-dev → azure-prod
+- target:
+    kind: Application
+    name: opencost
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/azure-prod/opencost-values.yaml
+
+# TODO: add patches for keycloak, grafana, secrets, enterprise-apps
+# when deploying to this cluster (these are deployment-specific, not cloud-specific)
--- a/infra/overlays/gcp-dev/kustomization.yaml
+++ b/infra/overlays/gcp-dev/kustomization.yaml
@@ -0,0 +1,35 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- ../../base
+
+patches:
+# Traefik: swap upc-dev → gcp-dev
+- target:
+    kind: Application
+    name: traefik
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/gcp-dev/traefik-values.yaml
+
+# Gitea: swap upc-dev → gcp-dev
+- target:
+    kind: Application
+    name: gitea
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/gcp-dev/gitea-values.yaml
+
+# OpenCost: swap upc-dev → gcp-dev
+- target:
+    kind: Application
+    name: opencost
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/gcp-dev/opencost-values.yaml
+
+# TODO: add patches for keycloak, grafana, secrets, enterprise-apps
+# when deploying to this cluster (these are deployment-specific, not cloud-specific)
--- a/infra/overlays/gcp-prod/kustomization.yaml
+++ b/infra/overlays/gcp-prod/kustomization.yaml
@@ -0,0 +1,35 @@
+apiVersion: kustomize.config.k8s.io/v1beta1
+kind: Kustomization
+resources:
+- ../../base
+
+patches:
+# Traefik: swap upc-dev → gcp-prod
+- target:
+    kind: Application
+    name: traefik
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/gcp-prod/traefik-values.yaml
+
+# Gitea: swap upc-dev → gcp-prod
+- target:
+    kind: Application
+    name: gitea
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/gcp-prod/gitea-values.yaml
+
+# OpenCost: swap upc-dev → gcp-prod
+- target:
+    kind: Application
+    name: opencost
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/gcp-prod/opencost-values.yaml
+
+# TODO: add patches for keycloak, grafana, secrets, enterprise-apps
+# when deploying to this cluster (these are deployment-specific, not cloud-specific)
--- a/infra/overlays/upc-prod/kustomization.yaml
+++ b/infra/overlays/upc-prod/kustomization.yaml
@@ -48,3 +48,21 @@ patches:
    - op: replace
      path: /spec/source/path
      value: apps/overlays/upc-prod
+
+# Gitea: swap upc-dev → upc-prod
+- target:
+    kind: Application
+    name: gitea
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/upc-prod/gitea-values.yaml
+
+# OpenCost: swap upc-dev → upc-prod
+- target:
+    kind: Application
+    name: opencost
+  patch: |
+    - op: replace
+      path: /spec/sources/0/helm/valueFiles/1
+      value: $values/infra/values/upc-prod/opencost-values.yaml
--- a/infra/values/aws-dev/gitea-values.yaml
+++ b/infra/values/aws-dev/gitea-values.yaml
@@ -0,0 +1,7 @@
+# AWS EBS gp3 storage class (requires EBS CSI driver)
+persistence:
+  storageClass: gp3
+postgresql:
+  primary:
+    persistence:
+      storageClass: gp3
--- a/infra/values/aws-dev/opencost-values.yaml
+++ b/infra/values/aws-dev/opencost-values.yaml
@@ -0,0 +1,13 @@
+# AWS native pricing via Cost and Usage Reports
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: aws
+    aws:
+      service_key_name: ""      # <- populate or use IRSA
+      service_key_secret: ""
+      spot_data_region: ""
+      spot_data_bucket: ""
+      spot_data_prefix: ""
+      account_id: ""
--- a/infra/values/aws-dev/traefik-values.yaml
+++ b/infra/values/aws-dev/traefik-values.yaml
@@ -0,0 +1,18 @@
+# AWS EKS — NLB with Proxy Protocol v2 for real client IPs
+service:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: "external"
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: "ip"
+    service.beta.kubernetes.io/aws-load-balancer-scheme: "internet-facing"
+    service.beta.kubernetes.io/aws-load-balancer-proxy-protocol: "*"
+ports:
+  web:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8"       # <- adjust to your VPC CIDR
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8"
+  websecure:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8"
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8"
--- a/infra/values/aws-prod/gitea-values.yaml
+++ b/infra/values/aws-prod/gitea-values.yaml
@@ -0,0 +1,7 @@
+# AWS EBS gp3 storage class (requires EBS CSI driver)
+persistence:
+  storageClass: gp3
+postgresql:
+  primary:
+    persistence:
+      storageClass: gp3
--- a/infra/values/aws-prod/opencost-values.yaml
+++ b/infra/values/aws-prod/opencost-values.yaml
@@ -0,0 +1,13 @@
+# AWS native pricing via Cost and Usage Reports
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: aws
+    aws:
+      service_key_name: ""      # <- populate or use IRSA
+      service_key_secret: ""
+      spot_data_region: ""
+      spot_data_bucket: ""
+      spot_data_prefix: ""
+      account_id: ""
--- a/infra/values/aws-prod/traefik-values.yaml
+++ b/infra/values/aws-prod/traefik-values.yaml
@@ -0,0 +1,18 @@
+# AWS EKS — NLB with Proxy Protocol v2 for real client IPs
+service:
+  annotations:
+    service.beta.kubernetes.io/aws-load-balancer-type: "external"
+    service.beta.kubernetes.io/aws-load-balancer-nlb-target-type: "ip"
+    service.beta.kubernetes.io/aws-load-balancer-scheme: "internet-facing"
+    service.beta.kubernetes.io/aws-load-balancer-proxy-protocol: "*"
+ports:
+  web:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8"       # <- adjust to your VPC CIDR
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8"
+  websecure:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8"
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8"
--- a/infra/values/azure-dev/gitea-values.yaml
+++ b/infra/values/azure-dev/gitea-values.yaml
@@ -0,0 +1,7 @@
+# Azure Managed Disk (Premium SSD via CSI driver)
+persistence:
+  storageClass: managed-csi-premium
+postgresql:
+  primary:
+    persistence:
+      storageClass: managed-csi-premium
--- a/infra/values/azure-dev/opencost-values.yaml
+++ b/infra/values/azure-dev/opencost-values.yaml
@@ -0,0 +1,11 @@
+# Azure native pricing via Billing API
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: azure
+    azure:
+      subscriptionID: ""          # <- populate
+      clientID: ""
+      clientSecret: ""
+      tenantID: ""
--- a/infra/values/azure-dev/traefik-values.yaml
+++ b/infra/values/azure-dev/traefik-values.yaml
@@ -0,0 +1,16 @@
+# Azure AKS — Standard Load Balancer
+# Note: Azure Standard LB does not support Proxy Protocol.
+# Use externalTrafficPolicy: Local on the Traefik service to preserve
+# client IPs, or deploy behind Azure Application Gateway.
+service:
+  annotations:
+    service.beta.kubernetes.io/azure-load-balancer-health-probe-request-path: "/ping"
+  spec:
+    externalTrafficPolicy: Local
+ports:
+  web:
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,168.63.129.16/32"    # <- VNet CIDR + Azure health probe
+  websecure:
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,168.63.129.16/32"
--- a/infra/values/azure-prod/gitea-values.yaml
+++ b/infra/values/azure-prod/gitea-values.yaml
@@ -0,0 +1,7 @@
+# Azure Managed Disk (Premium SSD via CSI driver)
+persistence:
+  storageClass: managed-csi-premium
+postgresql:
+  primary:
+    persistence:
+      storageClass: managed-csi-premium
--- a/infra/values/azure-prod/opencost-values.yaml
+++ b/infra/values/azure-prod/opencost-values.yaml
@@ -0,0 +1,11 @@
+# Azure native pricing via Billing API
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: azure
+    azure:
+      subscriptionID: ""          # <- populate
+      clientID: ""
+      clientSecret: ""
+      tenantID: ""
--- a/infra/values/azure-prod/traefik-values.yaml
+++ b/infra/values/azure-prod/traefik-values.yaml
@@ -0,0 +1,16 @@
+# Azure AKS — Standard Load Balancer
+# Note: Azure Standard LB does not support Proxy Protocol.
+# Use externalTrafficPolicy: Local on the Traefik service to preserve
+# client IPs, or deploy behind Azure Application Gateway.
+service:
+  annotations:
+    service.beta.kubernetes.io/azure-load-balancer-health-probe-request-path: "/ping"
+  spec:
+    externalTrafficPolicy: Local
+ports:
+  web:
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,168.63.129.16/32"    # <- VNet CIDR + Azure health probe
+  websecure:
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,168.63.129.16/32"
--- a/infra/values/base/gitea-values.yaml
+++ b/infra/values/base/gitea-values.yaml
@@ -127,7 +127,6 @@ persistence:
  size: 10Gi
  accessModes:
  - ReadWriteOnce
-  storageClass: upcloud-block-storage-maxiops

 # -- Recreate strategy to avoid Multi-Attach errors with RWO volumes
 strategy:
@@ -153,7 +152,6 @@ postgresql:
    persistence:
      enabled: true
      size: 8Gi
-      storageClass: upcloud-block-storage-maxiops
    resources:
      requests:
        cpu: 100m
--- a/infra/values/base/opencost-values.yaml
+++ b/infra/values/base/opencost-values.yaml
@@ -10,18 +10,8 @@ opencost:
        serviceName: prometheus-server
        namespaceName: monitoring
        port: 80
-    customPricing:
-      enabled: true
-      provider: custom
-      costModel:
-        description: "UpCloud 4-node cluster pricing"
-        CPU: "5.86"
-        RAM: "1.46"
-        GPU: "0"
-        storage: "0.34"
-        zoneNetworkEgress: "0"
-        regionNetworkEgress: "0"
-        internetNetworkEgress: "0"
+    # Cloud-specific pricing is in per-cluster value overrides
+    # (e.g. infra/values/upc-dev/opencost-values.yaml)
  ui:
    enabled: false
 service:
--- a/infra/values/gcp-dev/gitea-values.yaml
+++ b/infra/values/gcp-dev/gitea-values.yaml
@@ -0,0 +1,7 @@
+# GCP Persistent Disk (SSD via CSI driver)
+persistence:
+  storageClass: premium-rwo
+postgresql:
+  primary:
+    persistence:
+      storageClass: premium-rwo
--- a/infra/values/gcp-dev/opencost-values.yaml
+++ b/infra/values/gcp-dev/opencost-values.yaml
@@ -0,0 +1,9 @@
+# GCP native pricing via Cloud Billing API
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: gcp
+    gcp:
+      projectID: ""               # <- populate with your GCP project ID
+      key: ""                     # <- or use Workload Identity
--- a/infra/values/gcp-dev/traefik-values.yaml
+++ b/infra/values/gcp-dev/traefik-values.yaml
@@ -0,0 +1,15 @@
+# GCP GKE — External passthrough Network Load Balancer
+service:
+  annotations:
+    cloud.google.com/l4-rbs: "enabled"
+ports:
+  web:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"   # <- subnet CIDR + GCP health checks
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"
+  websecure:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"
--- a/infra/values/gcp-prod/gitea-values.yaml
+++ b/infra/values/gcp-prod/gitea-values.yaml
@@ -0,0 +1,7 @@
+# GCP Persistent Disk (SSD via CSI driver)
+persistence:
+  storageClass: premium-rwo
+postgresql:
+  primary:
+    persistence:
+      storageClass: premium-rwo
--- a/infra/values/gcp-prod/opencost-values.yaml
+++ b/infra/values/gcp-prod/opencost-values.yaml
@@ -0,0 +1,9 @@
+# GCP native pricing via Cloud Billing API
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: gcp
+    gcp:
+      projectID: ""               # <- populate with your GCP project ID
+      key: ""                     # <- or use Workload Identity
--- a/infra/values/gcp-prod/traefik-values.yaml
+++ b/infra/values/gcp-prod/traefik-values.yaml
@@ -0,0 +1,15 @@
+# GCP GKE — External passthrough Network Load Balancer
+service:
+  annotations:
+    cloud.google.com/l4-rbs: "enabled"
+ports:
+  web:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"   # <- subnet CIDR + GCP health checks
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"
+  websecure:
+    proxyProtocol:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"
+    forwardedHeaders:
+      trustedIPs: "10.0.0.0/8,35.191.0.0/16,130.211.0.0/22"
--- a/infra/values/upc-dev/gitea-values.yaml
+++ b/infra/values/upc-dev/gitea-values.yaml
@@ -0,0 +1,7 @@
+# UpCloud storage class for Gitea and its embedded PostgreSQL
+persistence:
+  storageClass: upcloud-block-storage-maxiops
+postgresql:
+  primary:
+    persistence:
+      storageClass: upcloud-block-storage-maxiops
--- a/infra/values/upc-dev/opencost-values.yaml
+++ b/infra/values/upc-dev/opencost-values.yaml
@@ -0,0 +1,15 @@
+# UpCloud custom pricing (no native OpenCost integration)
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: custom
+      costModel:
+        description: "UpCloud 4-node cluster pricing"
+        CPU: "5.86"
+        RAM: "1.46"
+        GPU: "0"
+        storage: "0.34"
+        zoneNetworkEgress: "0"
+        regionNetworkEgress: "0"
+        internetNetworkEgress: "0"
--- a/infra/values/upc-prod/gitea-values.yaml
+++ b/infra/values/upc-prod/gitea-values.yaml
@@ -0,0 +1,7 @@
+# UpCloud storage class for Gitea and its embedded PostgreSQL
+persistence:
+  storageClass: upcloud-block-storage-maxiops
+postgresql:
+  primary:
+    persistence:
+      storageClass: upcloud-block-storage-maxiops
--- a/infra/values/upc-prod/opencost-values.yaml
+++ b/infra/values/upc-prod/opencost-values.yaml
@@ -0,0 +1,15 @@
+# UpCloud custom pricing (no native OpenCost integration)
+opencost:
+  exporter:
+    customPricing:
+      enabled: true
+      provider: custom
+      costModel:
+        description: "UpCloud 4-node cluster pricing"
+        CPU: "5.86"
+        RAM: "1.46"
+        GPU: "0"
+        storage: "0.34"
+        zoneNetworkEgress: "0"
+        regionNetworkEgress: "0"
+        internetNetworkEgress: "0"
--- a/scripts/gitea-backup-aws.sh
+++ b/scripts/gitea-backup-aws.sh
@@ -0,0 +1,94 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# Gitea backup helper for AWS S3
+# Uses the gitea-backup-s3 secret in the gitea namespace
+# (same secret schema: S3_ENDPOINT, AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, S3_BUCKET)
+#
+# For AWS, S3_ENDPOINT is typically https://s3.<region>.amazonaws.com
+#
+# Usage:
+#   ./scripts/gitea-backup-aws.sh list                    # list all backups
+#   ./scripts/gitea-backup-aws.sh download <filename>     # download a backup to current dir
+#   ./scripts/gitea-backup-aws.sh download latest          # download the most recent backup
+
+NAMESPACE="gitea"
+SECRET="gitea-backup-s3"
+IMAGE="minio/mc:latest"
+POD_NAME="gitea-backup-helper"
+ALIAS_CMD='mc alias set s3 ${S3_ENDPOINT} ${AWS_ACCESS_KEY_ID} ${AWS_SECRET_ACCESS_KEY} > /dev/null'
+
+cleanup() {
+  kubectl -n "$NAMESPACE" delete pod "$POD_NAME" --ignore-not-found --grace-period=0 > /dev/null 2>&1 || true
+}
+
+mc_run() {
+  cleanup
+  kubectl -n "$NAMESPACE" run "$POD_NAME" --restart=Never \
+    --image="$IMAGE" \
+    --overrides="{
+      \"spec\":{\"containers\":[{
+        \"name\":\"$POD_NAME\",
+        \"image\":\"$IMAGE\",
+        \"env\":[{\"name\":\"HOME\",\"value\":\"/tmp\"}],
+        \"command\":[\"sh\",\"-c\",\"${ALIAS_CMD}; $1\"],
+        \"envFrom\":[{\"secretRef\":{\"name\":\"$SECRET\"}}]
+      }]}
+    }" > /dev/null 2>&1
+
+  kubectl -n "$NAMESPACE" wait --for=jsonpath='{.status.phase}'=Succeeded "pod/$POD_NAME" --timeout=120s > /dev/null 2>&1
+  kubectl -n "$NAMESPACE" logs "$POD_NAME"
+  cleanup
+}
+
+case "${1:-help}" in
+  list)
+    echo "Listing backups..."
+    mc_run 'mc ls s3/${S3_BUCKET}/'
+    ;;
+
+  download)
+    FILE="${2:?Usage: $0 download <filename|latest>}"
+
+    if [ "$FILE" = "latest" ]; then
+      echo "Finding latest backup..."
+      FILE=$(mc_run 'mc ls s3/${S3_BUCKET}/' | sort | tail -1 | awk '{print $NF}' | tr -d '[:space:]')
+      if [ -z "$FILE" ]; then
+        echo "No backups found."
+        exit 1
+      fi
+      echo "Latest: $FILE"
+    fi
+
+    echo "Downloading $FILE..."
+    cleanup
+    kubectl -n "$NAMESPACE" run "$POD_NAME" --restart=Never \
+      --image="$IMAGE" \
+      --overrides="{
+        \"spec\":{\"containers\":[{
+          \"name\":\"$POD_NAME\",
+          \"image\":\"$IMAGE\",
+          \"env\":[{\"name\":\"HOME\",\"value\":\"/tmp\"}],
+          \"command\":[\"sh\",\"-c\",\"sleep 300\"],
+          \"envFrom\":[{\"secretRef\":{\"name\":\"$SECRET\"}}]
+        }]}
+      }" > /dev/null 2>&1
+
+    kubectl -n "$NAMESPACE" wait --for=condition=Ready "pod/$POD_NAME" --timeout=60s > /dev/null 2>&1
+
+    echo "Saving to ./$FILE ..."
+    kubectl -n "$NAMESPACE" exec "$POD_NAME" -- sh -c "${ALIAS_CMD} && mc cat s3/\${S3_BUCKET}/$FILE" > "./$FILE"
+    cleanup
+
+    echo "Downloaded: ./$FILE"
+    ;;
+
+  *)
+    echo "Gitea backup helper (AWS S3)"
+    echo ""
+    echo "Usage:"
+    echo "  $0 list                    List all backups in S3"
+    echo "  $0 download <filename>     Download a specific backup"
+    echo "  $0 download latest         Download the most recent backup"
+    ;;
+esac
--- a/scripts/gitea-backup-azure.sh
+++ b/scripts/gitea-backup-azure.sh
@@ -0,0 +1,100 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# Gitea backup helper for Azure Blob Storage
+# Uses the gitea-backup-azure secret in the gitea namespace
+# Required secret keys:
+#   AZURE_STORAGE_ACCOUNT   — storage account name
+#   AZURE_STORAGE_KEY       — storage account key
+#   AZURE_CONTAINER         — blob container name
+#
+# Usage:
+#   ./scripts/gitea-backup-azure.sh list                    # list all backups
+#   ./scripts/gitea-backup-azure.sh download <filename>     # download a backup
+#   ./scripts/gitea-backup-azure.sh download latest          # download the most recent backup
+
+NAMESPACE="gitea"
+SECRET="gitea-backup-azure"
+IMAGE="mcr.microsoft.com/azure-cli:latest"
+POD_NAME="gitea-backup-helper"
+
+cleanup() {
+  kubectl -n "$NAMESPACE" delete pod "$POD_NAME" --ignore-not-found --grace-period=0 > /dev/null 2>&1 || true
+}
+
+az_run() {
+  cleanup
+  kubectl -n "$NAMESPACE" run "$POD_NAME" --restart=Never \
+    --image="$IMAGE" \
+    --overrides="{
+      \"spec\":{\"containers\":[{
+        \"name\":\"$POD_NAME\",
+        \"image\":\"$IMAGE\",
+        \"env\":[{\"name\":\"HOME\",\"value\":\"/tmp\"}],
+        \"command\":[\"sh\",\"-c\",\"$1\"],
+        \"envFrom\":[{\"secretRef\":{\"name\":\"$SECRET\"}}]
+      }]}
+    }" > /dev/null 2>&1
+
+  kubectl -n "$NAMESPACE" wait --for=jsonpath='{.status.phase}'=Succeeded "pod/$POD_NAME" --timeout=120s > /dev/null 2>&1
+  kubectl -n "$NAMESPACE" logs "$POD_NAME"
+  cleanup
+}
+
+case "${1:-help}" in
+  list)
+    echo "Listing backups..."
+    az_run 'az storage blob list --account-name ${AZURE_STORAGE_ACCOUNT} --account-key ${AZURE_STORAGE_KEY} --container-name ${AZURE_CONTAINER} --output table --query "[].{Name:name, Size:properties.contentLength, Modified:properties.lastModified}"'
+    ;;
+
+  download)
+    FILE="${2:?Usage: $0 download <filename|latest>}"
+
+    if [ "$FILE" = "latest" ]; then
+      echo "Finding latest backup..."
+      FILE=$(az_run 'az storage blob list --account-name ${AZURE_STORAGE_ACCOUNT} --account-key ${AZURE_STORAGE_KEY} --container-name ${AZURE_CONTAINER} --query "sort_by([], &properties.lastModified)[-1].name" -o tsv' | tr -d '[:space:]')
+      if [ -z "$FILE" ]; then
+        echo "No backups found."
+        exit 1
+      fi
+      echo "Latest: $FILE"
+    fi
+
+    echo "Downloading $FILE..."
+    cleanup
+    kubectl -n "$NAMESPACE" run "$POD_NAME" --restart=Never \
+      --image="$IMAGE" \
+      --overrides="{
+        \"spec\":{\"containers\":[{
+          \"name\":\"$POD_NAME\",
+          \"image\":\"$IMAGE\",
+          \"env\":[{\"name\":\"HOME\",\"value\":\"/tmp\"}],
+          \"command\":[\"sh\",\"-c\",\"sleep 300\"],
+          \"envFrom\":[{\"secretRef\":{\"name\":\"$SECRET\"}}]
+        }]}
+      }" > /dev/null 2>&1
+
+    kubectl -n "$NAMESPACE" wait --for=condition=Ready "pod/$POD_NAME" --timeout=60s > /dev/null 2>&1
+
+    echo "Saving to ./$FILE ..."
+    kubectl -n "$NAMESPACE" exec "$POD_NAME" -- \
+      az storage blob download \
+        --account-name "\${AZURE_STORAGE_ACCOUNT}" \
+        --account-key "\${AZURE_STORAGE_KEY}" \
+        --container-name "\${AZURE_CONTAINER}" \
+        --name "$FILE" \
+        --file /dev/stdout 2>/dev/null > "./$FILE"
+    cleanup
+
+    echo "Downloaded: ./$FILE"
+    ;;
+
+  *)
+    echo "Gitea backup helper (Azure Blob Storage)"
+    echo ""
+    echo "Usage:"
+    echo "  $0 list                    List all backups in Azure Blob"
+    echo "  $0 download <filename>     Download a specific backup"
+    echo "  $0 download latest         Download the most recent backup"
+    ;;
+esac
--- a/scripts/gitea-backup-gcp.sh
+++ b/scripts/gitea-backup-gcp.sh
@@ -0,0 +1,95 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# Gitea backup helper for Google Cloud Storage
+# Uses the gitea-backup-gcs secret in the gitea namespace
+# Required secret keys:
+#   GCS_BUCKET              — bucket name (without gs:// prefix)
+#   GOOGLE_APPLICATION_CREDENTIALS_JSON — service account key JSON
+#     (alternatively, use Workload Identity and omit the key)
+#
+# Usage:
+#   ./scripts/gitea-backup-gcp.sh list                    # list all backups
+#   ./scripts/gitea-backup-gcp.sh download <filename>     # download a backup
+#   ./scripts/gitea-backup-gcp.sh download latest          # download the most recent backup
+
+NAMESPACE="gitea"
+SECRET="gitea-backup-gcs"
+IMAGE="gcr.io/google.com/cloudsdktool/google-cloud-cli:slim"
+POD_NAME="gitea-backup-helper"
+AUTH_CMD='if [ -n "${GOOGLE_APPLICATION_CREDENTIALS_JSON:-}" ]; then echo "${GOOGLE_APPLICATION_CREDENTIALS_JSON}" > /tmp/gcs-key.json && gcloud auth activate-service-account --key-file=/tmp/gcs-key.json > /dev/null 2>&1; fi'
+
+cleanup() {
+  kubectl -n "$NAMESPACE" delete pod "$POD_NAME" --ignore-not-found --grace-period=0 > /dev/null 2>&1 || true
+}
+
+gcs_run() {
+  cleanup
+  kubectl -n "$NAMESPACE" run "$POD_NAME" --restart=Never \
+    --image="$IMAGE" \
+    --overrides="{
+      \"spec\":{\"containers\":[{
+        \"name\":\"$POD_NAME\",
+        \"image\":\"$IMAGE\",
+        \"env\":[{\"name\":\"HOME\",\"value\":\"/tmp\"}],
+        \"command\":[\"sh\",\"-c\",\"${AUTH_CMD}; $1\"],
+        \"envFrom\":[{\"secretRef\":{\"name\":\"$SECRET\"}}]
+      }]}
+    }" > /dev/null 2>&1
+
+  kubectl -n "$NAMESPACE" wait --for=jsonpath='{.status.phase}'=Succeeded "pod/$POD_NAME" --timeout=120s > /dev/null 2>&1
+  kubectl -n "$NAMESPACE" logs "$POD_NAME"
+  cleanup
+}
+
+case "${1:-help}" in
+  list)
+    echo "Listing backups..."
+    gcs_run 'gsutil ls -l gs://${GCS_BUCKET}/'
+    ;;
+
+  download)
+    FILE="${2:?Usage: $0 download <filename|latest>}"
+
+    if [ "$FILE" = "latest" ]; then
+      echo "Finding latest backup..."
+      FILE=$(gcs_run 'gsutil ls gs://${GCS_BUCKET}/' | grep -v '^$' | grep -v 'TOTAL' | sort | tail -1 | xargs -I{} basename {} | tr -d '[:space:]')
+      if [ -z "$FILE" ]; then
+        echo "No backups found."
+        exit 1
+      fi
+      echo "Latest: $FILE"
+    fi
+
+    echo "Downloading $FILE..."
+    cleanup
+    kubectl -n "$NAMESPACE" run "$POD_NAME" --restart=Never \
+      --image="$IMAGE" \
+      --overrides="{
+        \"spec\":{\"containers\":[{
+          \"name\":\"$POD_NAME\",
+          \"image\":\"$IMAGE\",
+          \"env\":[{\"name\":\"HOME\",\"value\":\"/tmp\"}],
+          \"command\":[\"sh\",\"-c\",\"sleep 300\"],
+          \"envFrom\":[{\"secretRef\":{\"name\":\"$SECRET\"}}]
+        }]}
+      }" > /dev/null 2>&1
+
+    kubectl -n "$NAMESPACE" wait --for=condition=Ready "pod/$POD_NAME" --timeout=60s > /dev/null 2>&1
+
+    echo "Saving to ./$FILE ..."
+    kubectl -n "$NAMESPACE" exec "$POD_NAME" -- sh -c "${AUTH_CMD} && gsutil cat gs://\${GCS_BUCKET}/$FILE" > "./$FILE"
+    cleanup
+
+    echo "Downloaded: ./$FILE"
+    ;;
+
+  *)
+    echo "Gitea backup helper (Google Cloud Storage)"
+    echo ""
+    echo "Usage:"
+    echo "  $0 list                    List all backups in GCS"
+    echo "  $0 download <filename>     Download a specific backup"
+    echo "  $0 download latest         Download the most recent backup"
+    ;;
+esac