droplet

deploy-local.ps1

@@ -0,0 +1,91 @@
+# AntiAtropos Local Cluster Deploy
+# Deploys workloads, Prometheus, and Grafana on the Kind cluster.
+# Grafana port-forward starts automatically at the end.
+
+param(
+    [switch]$SkipPortForward,
+    [int]$GrafanaPort = 3000
+)
+
+Write-Host "=== AntiAtropos Local Deploy ===" -ForegroundColor Cyan
+Write-Host ""
+
+# --- 1. Check cluster ---
+Write-Host "[1/5] Checking Kind cluster..." -ForegroundColor Yellow
+$cluster = kubectl config current-context 2>$null
+if ($cluster -notmatch "antiatropos") {
+    Write-Host "WARNING: Current context is '$cluster', expected 'kind-antiatropos-local'. Proceed anyway? [Y/n]"
+    $r = Read-Host
+    if ($r -eq 'n') { exit 1 }
+}
+
+# --- 2. Deploy workload pods ---
+Write-Host "[2/5] Deploying workload pods..." -ForegroundColor Yellow
+kubectl create ns prod-sre 2>&1 | Out-Null
+kubectl create ns monitoring 2>&1 | Out-Null
+kubectl apply -f "$PSScriptRoot\deploy\local-laptop.yaml"
+Write-Host "  Waiting for workloads to be ready..."
+kubectl wait --for=condition=ready pod -l app --all -n prod-sre --timeout=120s 2>$null
+Write-Host "  Workloads ready."
+
+# --- 3. Deploy Prometheus ---
+Write-Host "[3/5] Deploying Prometheus..." -ForegroundColor Yellow
+$promRelease = helm list -n monitoring -q 2>$null | Select-String "prometheus"
+if ($promRelease) {
+    helm upgrade prometheus prometheus-community/prometheus -n monitoring -f "$PSScriptRoot\deploy\prometheus-helm-values.yaml"
+} else {
+    helm install prometheus prometheus-community/prometheus -n monitoring -f "$PSScriptRoot\deploy\prometheus-helm-values.yaml"
+}
+Write-Host "  Waiting for Prometheus server..."
+kubectl wait --for=condition=ready pod -l "app.kubernetes.io/name=prometheus" -n monitoring --timeout=120s 2>$null
+Write-Host "  Prometheus ready."
+
+# --- 4. Deploy Grafana ---
+Write-Host "[4/5] Deploying Grafana..." -ForegroundColor Yellow
+# Update dashboard ConfigMap
+kubectl delete configmap grafana-dashboards -n monitoring 2>$null
+kubectl create configmap grafana-dashboards -n monitoring --from-file="$PSScriptRoot\deploy\grafana\provisioning\dashboards\json\"
+
+$grafRelease = helm list -n monitoring -q 2>$null | Select-String "grafana"
+if ($grafRelease) {
+    helm upgrade grafana grafana/grafana -n monitoring -f "$PSScriptRoot\deploy\grafana-helm-values.yaml"
+} else {
+    helm install grafana grafana/grafana -n monitoring -f "$PSScriptRoot\deploy\grafana-helm-values.yaml"
+}
+Write-Host "  Waiting for Grafana..."
+kubectl wait --for=condition=ready pod -l "app.kubernetes.io/name=grafana" -n monitoring --timeout=120s 2>$null
+Write-Host "  Grafana ready."
+
+# --- 5. Start Grafana port-forward ---
+Write-Host "[5/5] Grafana port-forward..." -ForegroundColor Yellow
+if (-not $SkipPortForward) {
+    # Kill any existing port-forward on the same port
+    $existing = Get-NetTCPConnection -LocalPort $GrafanaPort -ErrorAction SilentlyContinue 2>$null
+    if ($existing) {
+        $pid = $existing.OwningProcess
+        Stop-Process -Id $pid -Force -ErrorAction SilentlyContinue 2>$null
+        Start-Sleep -Seconds 1
+    }
+
+    Write-Host "  Starting port-forward on localhost:$GrafanaPort..."
+    $proc = Start-Process -PassThru -NoNewWindow kubectl -ArgumentList "port-forward","-n","monitoring","svc/grafana","${GrafanaPort}:80"
+
+    Start-Sleep -Seconds 2
+    # Verify the port-forward is alive
+    try {
+        $null = Invoke-WebRequest -Uri "http://localhost:$GrafanaPort/api/health" -UseBasicParsing -TimeoutSec 5
+        Write-Host ""
+        Write-Host "=== Deploy Complete ===" -ForegroundColor Green
+        Write-Host "  Grafana:  http://localhost:$GrafanaPort  (admin / antiatropos)"
+        Write-Host "  Dashboards: AntiAtropos Overview, AntiAtropos Live Control Plane"
+        Write-Host "  Port-forward PID: $($proc.Id)"
+        Write-Host ""
+        Write-Host "To stop port-forward: Stop-Process -Id $($proc.Id)"
+    } catch {
+        Write-Host "WARNING: Port-forward started but Grafana not reachable yet. Try: kubectl port-forward -n monitoring svc/grafana ${GrafanaPort}:80"
+    }
+} else {
+    Write-Host ""
+    Write-Host "=== Deploy Complete ===" -ForegroundColor Green
+    Write-Host "  To access Grafana: kubectl port-forward -n monitoring svc/grafana ${GrafanaPort}:80"
+}

deploy/do/README.md

@@ -0,0 +1,50 @@
+# DigitalOcean Droplet one-shot deploy
+
+This deploy flow is for a single Ubuntu Droplet running:
+- k3s (single-node Kubernetes)
+- AntiAtropos sample workloads (`prod-sre`)
+- Prometheus + Grafana (`monitoring`)
+- FastAPI control server (`antiatropos-fastapi` systemd service)
+
+## Run
+
+From repository root on the Droplet:
+
+```bash
+sudo bash deploy/do/deploy-droplet-one-shot.sh
+```
+
+Optional overrides:
+
+```bash
+sudo REPO_DIR=/opt/AntiAtropos FASTAPI_PORT=8010 MAX_REPLICAS=200 bash deploy/do/deploy-droplet-one-shot.sh
+```
+
+## What the script configures
+
+- k3s kubelet with `max-pods=250`
+- Env file at `.env.droplet` with:
+  - `ANTIATROPOS_ENV_MODE=live`
+  - `KUBECONFIG=/etc/rancher/k3s/k3s.yaml`
+  - `ANTIATROPOS_WORKLOAD_MAP` for `node-0`..`node-4`
+- Systemd service:
+  - Name: `antiatropos-fastapi`
+  - Exec: `uvicorn server.app:app --host 0.0.0.0 --port 8000`
+
+## Verify
+
+```bash
+systemctl status antiatropos-fastapi --no-pager
+curl http://127.0.0.1:8000/config/runtime
+kubectl get deploy -n prod-sre
+kubectl get pods -n monitoring
+kubectl -n monitoring port-forward svc/grafana 3000:80
+```
+
+## Agent call example
+
+```bash
+curl -X POST http://127.0.0.1:8000/step \
+  -H "Content-Type: application/json" \
+  -d '{"action_type":"SCALE_UP","target_node_id":"node-3","parameter":0.6}'
+```

deploy/do/deploy-droplet-one-shot.sh

@@ -0,0 +1,150 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+# One-shot deploy for a single DigitalOcean Droplet:
+# - Installs k3s with kubelet max-pods=250
+# - Deploys workloads + Prometheus + Grafana
+# - Creates env file for live Kubernetes scaling
+# - Starts FastAPI server via systemd (antiatropos-fastapi)
+
+if [[ "${EUID}" -ne 0 ]]; then
+  echo "Run as root: sudo bash deploy/do/deploy-droplet-one-shot.sh"
+  exit 1
+fi
+
+REPO_DIR="${REPO_DIR:-$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)}"
+KUBECONFIG_PATH="${KUBECONFIG_PATH:-/etc/rancher/k3s/k3s.yaml}"
+FASTAPI_PORT="${FASTAPI_PORT:-8000}"
+FASTAPI_HOST="${FASTAPI_HOST:-0.0.0.0}"
+K8S_NAMESPACE="${K8S_NAMESPACE:-prod-sre}"
+MONITORING_NAMESPACE="${MONITORING_NAMESPACE:-monitoring}"
+PY_VENV_DIR="${PY_VENV_DIR:-${REPO_DIR}/.venv-droplet}"
+ENV_FILE="${ENV_FILE:-${REPO_DIR}/.env.droplet}"
+MIN_REPLICAS="${MIN_REPLICAS:-1}"
+MAX_REPLICAS="${MAX_REPLICAS:-250}"
+SCALE_STEP="${SCALE_STEP:-3}"
+WORKLOAD_MAP="${WORKLOAD_MAP:-{\"node-0\":{\"deployment\":\"payments\",\"namespace\":\"prod-sre\"},\"node-1\":{\"deployment\":\"checkout\",\"namespace\":\"prod-sre\"},\"node-2\":{\"deployment\":\"catalog\",\"namespace\":\"prod-sre\"},\"node-3\":{\"deployment\":\"cart\",\"namespace\":\"prod-sre\"},\"node-4\":{\"deployment\":\"auth\",\"namespace\":\"prod-sre\"}}}"
+
+echo "=== AntiAtropos Droplet One-Shot Deploy ==="
+echo "Repo:        ${REPO_DIR}"
+echo "Kubeconfig:  ${KUBECONFIG_PATH}"
+echo "FastAPI:     ${FASTAPI_HOST}:${FASTAPI_PORT}"
+echo ""
+
+if [[ ! -f "${REPO_DIR}/deploy/local-laptop.yaml" ]]; then
+  echo "ERROR: deploy/local-laptop.yaml not found. Run from AntiAtropos checkout."
+  exit 1
+fi
+
+export DEBIAN_FRONTEND=noninteractive
+apt-get update
+apt-get install -y curl ca-certificates gnupg lsb-release python3 python3-venv python3-pip
+
+if ! command -v kubectl >/dev/null 2>&1; then
+  echo "Installing k3s (includes kubectl)..."
+  curl -sfL https://get.k3s.io | sh -s - --write-kubeconfig-mode 644 --kubelet-arg=max-pods=250
+else
+  echo "k3s/kubectl already present; skipping k3s install."
+fi
+
+if ! command -v helm >/dev/null 2>&1; then
+  echo "Installing Helm..."
+  curl -fsSL https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3 | bash
+fi
+
+export KUBECONFIG="${KUBECONFIG_PATH}"
+
+echo "Waiting for Kubernetes node to be Ready..."
+kubectl wait --for=condition=Ready node --all --timeout=180s
+
+kubectl create ns "${K8S_NAMESPACE}" >/dev/null 2>&1 || true
+kubectl create ns "${MONITORING_NAMESPACE}" >/dev/null 2>&1 || true
+
+echo "Deploying AntiAtropos workloads..."
+kubectl apply -f "${REPO_DIR}/deploy/local-laptop.yaml"
+
+echo "Installing/upgrading Prometheus + Grafana..."
+helm repo add prometheus-community https://prometheus-community.github.io/helm-charts >/dev/null 2>&1 || true
+helm repo add grafana https://grafana.github.io/helm-charts >/dev/null 2>&1 || true
+helm repo update
+
+helm upgrade --install prometheus prometheus-community/prometheus \
+  -n "${MONITORING_NAMESPACE}" \
+  -f "${REPO_DIR}/deploy/prometheus-helm-values.yaml"
+
+if [[ -d "${REPO_DIR}/deploy/grafana/provisioning/dashboards/json" ]]; then
+  kubectl delete configmap grafana-dashboards -n "${MONITORING_NAMESPACE}" >/dev/null 2>&1 || true
+  kubectl create configmap grafana-dashboards \
+    -n "${MONITORING_NAMESPACE}" \
+    --from-file="${REPO_DIR}/deploy/grafana/provisioning/dashboards/json/"
+fi
+
+helm upgrade --install grafana grafana/grafana \
+  -n "${MONITORING_NAMESPACE}" \
+  -f "${REPO_DIR}/deploy/grafana-helm-values.yaml"
+
+if [[ ! -f "${ENV_FILE}" ]]; then
+  cat > "${ENV_FILE}" <<EOF
+ANTIATROPOS_ENV_MODE=live
+KUBECONFIG=/etc/rancher/k3s/k3s.yaml
+ANTIATROPOS_K8S_NAMESPACE=prod-sre
+ANTIATROPOS_MIN_REPLICAS=${MIN_REPLICAS}
+ANTIATROPOS_MAX_REPLICAS=${MAX_REPLICAS}
+ANTIATROPOS_SCALE_STEP=${SCALE_STEP}
+ANTIATROPOS_WORKLOAD_MAP=${WORKLOAD_MAP}
+EOF
+  echo "Created ${ENV_FILE}"
+else
+  echo "Using existing ${ENV_FILE}"
+fi
+
+echo "Preparing Python environment..."
+python3 -m venv "${PY_VENV_DIR}"
+"${PY_VENV_DIR}/bin/python" -m pip install --upgrade pip
+"${PY_VENV_DIR}/bin/pip" install -r "${REPO_DIR}/server/requirements.txt"
+"${PY_VENV_DIR}/bin/pip" install -e "${REPO_DIR}"
+
+cat > /etc/systemd/system/antiatropos-fastapi.service <<EOF
+[Unit]
+Description=AntiAtropos FastAPI Server
+After=network-online.target k3s.service
+Wants=network-online.target
+
+[Service]
+Type=simple
+User=root
+WorkingDirectory=${REPO_DIR}
+EnvironmentFile=${ENV_FILE}
+ExecStart=${PY_VENV_DIR}/bin/uvicorn server.app:app --host ${FASTAPI_HOST} --port ${FASTAPI_PORT}
+Restart=always
+RestartSec=3
+
+[Install]
+WantedBy=multi-user.target
+EOF
+
+systemctl daemon-reload
+systemctl enable --now antiatropos-fastapi
+
+echo ""
+echo "Waiting for app readiness..."
+for _ in {1..30}; do
+  if curl -fsS "http://127.0.0.1:${FASTAPI_PORT}/config/runtime" >/dev/null 2>&1; then
+    break
+  fi
+  sleep 2
+done
+
+echo ""
+echo "=== Deploy Complete ==="
+echo "FastAPI runtime: http://127.0.0.1:${FASTAPI_PORT}/config/runtime"
+echo "FastAPI health:  http://127.0.0.1:${FASTAPI_PORT}/state"
+echo "Prometheus svc:  kubectl -n ${MONITORING_NAMESPACE} get svc prometheus-server"
+echo "Grafana access:  kubectl -n ${MONITORING_NAMESPACE} port-forward svc/grafana 3000:80"
+echo ""
+echo "Service status command:"
+echo "  systemctl status antiatropos-fastapi --no-pager"
+echo ""
+echo "If needed, edit env and restart:"
+echo "  ${ENV_FILE}"
+echo "  systemctl restart antiatropos-fastapi"

deploy/grafana-datasource-local.yaml

@@ -0,0 +1,11 @@
+# Grafana datasource provisioning - points to in-cluster Prometheus
+apiVersion: 1
+
+datasources:
+  - name: Prometheus
+    uid: PBFA97CFB590B2093
+    type: prometheus
+    access: proxy
+    url: http://prometheus-server.monitoring.svc.cluster.local
+    isDefault: true
+    editable: true

deploy/grafana-helm-values.yaml

@@ -0,0 +1,46 @@
+# Grafana self-hosted on Kind - Simplified dashboard + datasource setup
+
+adminUser: admin
+adminPassword: antiatropos
+
+service:
+  type: ClusterIP
+  port: 80
+
+persistence:
+  enabled: false
+
+# Datasource provisioning - mount as separate file
+datasources:
+  datasources.yaml:
+    apiVersion: 1
+    datasources:
+      - name: Prometheus
+        uid: PBFA97CFB590B2093
+        type: prometheus
+        access: proxy
+        url: http://prometheus-server.monitoring.svc.cluster.local
+        isDefault: true
+        editable: true
+
+# Dashboard provider config
+dashboardProviders:
+  dashboardproviders.yaml:
+    apiVersion: 1
+    providers:
+      - name: AntiAtropos
+        orgId: 1
+        folder: AntiAtropos
+        type: file
+        disableDeletion: false
+        editable: true
+        updateIntervalSeconds: 30
+        options:
+          path: /var/lib/grafana/dashboards/antiatropos
+
+# Mount dashboard JSONs from ConfigMap
+extraConfigmapMounts:
+  - name: grafana-dashboards
+    configMap: grafana-dashboards
+    mountPath: /var/lib/grafana/dashboards/antiatropos
+    readOnly: true

deploy/local-laptop.yaml

@@ -20,7 +20,7 @@ spec:
       annotations:
         prometheus.io/scrape: "true"
         prometheus.io/port: "8080"
-        prometheus.io/path: "/metrics"
+        prometheus.io/path: "/metrics.txt"
     spec:
       containers:
       - name: auth
@@ -71,7 +71,7 @@ spec:
               echo "http_request_duration_seconds_bucket{node_id=\"${NODE_ID}\",le=\"+Inf\"} ${req}"
               echo "http_request_duration_seconds_count{node_id=\"${NODE_ID}\"} ${req}"
               echo "http_request_duration_seconds_sum{node_id=\"${NODE_ID}\"} ${lat_sum}"
-            } > /www/metrics
+            } > /www/metrics.txt
             sleep 2
           done
 ---
@@ -92,7 +92,7 @@ spec:
       annotations:
         prometheus.io/scrape: "true"
         prometheus.io/port: "8080"
-        prometheus.io/path: "/metrics"
+        prometheus.io/path: "/metrics.txt"
     spec:
       containers:
       - name: cart
@@ -143,7 +143,7 @@ spec:
               echo "http_request_duration_seconds_bucket{node_id=\"${NODE_ID}\",le=\"+Inf\"} ${req}"
               echo "http_request_duration_seconds_count{node_id=\"${NODE_ID}\"} ${req}"
               echo "http_request_duration_seconds_sum{node_id=\"${NODE_ID}\"} ${lat_sum}"
-            } > /www/metrics
+            } > /www/metrics.txt
             sleep 2
           done
 ---
@@ -164,7 +164,7 @@ spec:
       annotations:
         prometheus.io/scrape: "true"
         prometheus.io/port: "8080"
-        prometheus.io/path: "/metrics"
+        prometheus.io/path: "/metrics.txt"
     spec:
       containers:
       - name: catalog
@@ -215,7 +215,7 @@ spec:
               echo "http_request_duration_seconds_bucket{node_id=\"${NODE_ID}\",le=\"+Inf\"} ${req}"
               echo "http_request_duration_seconds_count{node_id=\"${NODE_ID}\"} ${req}"
               echo "http_request_duration_seconds_sum{node_id=\"${NODE_ID}\"} ${lat_sum}"
-            } > /www/metrics
+            } > /www/metrics.txt
             sleep 2
           done
 ---
@@ -236,7 +236,7 @@ spec:
       annotations:
         prometheus.io/scrape: "true"
         prometheus.io/port: "8080"
-        prometheus.io/path: "/metrics"
+        prometheus.io/path: "/metrics.txt"
     spec:
       containers:
       - name: checkout
@@ -287,7 +287,7 @@ spec:
               echo "http_request_duration_seconds_bucket{node_id=\"${NODE_ID}\",le=\"+Inf\"} ${req}"
               echo "http_request_duration_seconds_count{node_id=\"${NODE_ID}\"} ${req}"
               echo "http_request_duration_seconds_sum{node_id=\"${NODE_ID}\"} ${lat_sum}"
-            } > /www/metrics
+            } > /www/metrics.txt
             sleep 2
           done
 ---
@@ -308,7 +308,7 @@ spec:
       annotations:
         prometheus.io/scrape: "true"
         prometheus.io/port: "8080"
-        prometheus.io/path: "/metrics"
+        prometheus.io/path: "/metrics.txt"
     spec:
       containers:
       - name: payments
@@ -359,7 +359,7 @@ spec:
               echo "http_request_duration_seconds_bucket{node_id=\"${NODE_ID}\",le=\"+Inf\"} ${req}"
               echo "http_request_duration_seconds_count{node_id=\"${NODE_ID}\"} ${req}"
               echo "http_request_duration_seconds_sum{node_id=\"${NODE_ID}\"} ${lat_sum}"
-            } > /www/metrics
+            } > /www/metrics.txt
             sleep 2
           done
 

deploy/prometheus-helm-values.yaml

@@ -0,0 +1,13 @@
+server:
+  global:
+    scrape_interval: 15s
+    evaluation_interval: 15s
+
+extraScrapeConfigs: |
+  - job_name: antiatropos-fastapi
+    scrape_protocols:
+      - PrometheusText1.0.0
+      - PrometheusText0.0.4
+    metrics_path: /metrics
+    static_configs:
+      - targets: ['host.docker.internal:8000']

deploy/prometheus-local.yml

@@ -0,0 +1,35 @@
+global:
+  scrape_interval: 15s
+  evaluation_interval: 15s
+  scrape_protocols:
+    - OpenMetricsText1.0.0
+    - OpenMetricsText0.0.1
+    - PrometheusText1.0.0
+    - PrometheusText0.0.4
+
+scrape_configs:
+  - job_name: kubernetes-pods
+    scrape_protocols:
+      - PrometheusText1.0.0
+      - PrometheusText0.0.4
+    kubernetes_sd_configs:
+      - role: pod
+    relabel_configs:
+      - source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_scrape]
+        action: keep
+        regex: true
+      - source_labels: [__meta_kubernetes_pod_annotation_prometheus_io_path]
+        action: replace
+        target_label: __metrics_path__
+        regex: (.+)
+      - source_labels: [__address__, __meta_kubernetes_pod_annotation_prometheus_io_port]
+        action: replace
+        regex: ([^:]+)(?::\d+)?;(\d+)
+        replacement: $1:$2
+        target_label: __address__
+      - source_labels: [__meta_kubernetes_namespace]
+        target_label: namespace
+      - source_labels: [__meta_kubernetes_pod_name]
+        target_label: pod
+      - source_labels: [__meta_kubernetes_pod_label_app]
+        target_label: app

simulator.py

@@ -64,8 +64,11 @@ CRITICAL_NODES: list[str] = ["node-0", "node-1", "node-2"]
 
 # VIP / business-critical node weights.
 # node-0 is the payment portal, so its queue growth or failure matters more.
+# Reduced from 4.0 → 2.0 to prevent reward gradient from creating
+# a local optimum where the agent only scales node-0.
+# At 2×, node-0 is still prioritized but other nodes remain viable targets.
 VIP_NODE_WEIGHTS: dict[str, float] = {
-    "node-0": 4.0,
+    "node-0": 2.0,
 }
 
 

start-grafana.ps1

@@ -1,12 +1,15 @@
-docker stop antiatropos-grafana 2>$null
-docker rm antiatropos-grafana 2>$null
+# Start Grafana port-forward from Kind cluster
+# Run 'deploy-local.ps1' first to ensure Grafana is deployed.
 
-Write-Host "Starting local Grafana (datasource -> host.docker.internal:9090)..."
+$port = 3000
 
-docker run -d --name antiatropos-grafana -p 3000:3000 `
-  -v "$PWD\deploy\grafana\provisioning:/etc/grafana/provisioning:ro" `
-  -e GF_AUTH_ANONYMOUS_ENABLED=true `
-  -e GF_AUTH_ANONYMOUS_ORG_ROLE=Admin `
-  grafana/grafana:latest | Out-Null
+Write-Host "Starting Grafana port-forward on localhost:$port..."
 
-Write-Host "Grafana is running at http://localhost:3000"
+# Kill any existing on that port
+$existing = Get-NetTCPConnection -LocalPort $port -ErrorAction SilentlyContinue
+if ($existing) {
+    Stop-Process -Id $existing.OwningProcess -Force -ErrorAction SilentlyContinue
+    Start-Sleep -Seconds 1
+}
+
+kubectl port-forward -n monitoring svc/grafana ${port}:80

teardown-local.ps1

@@ -0,0 +1,28 @@
+# AntiAtropos Local Cluster Teardown
+# Removes workloads, Prometheus, and Grafana. Stops port-forwards.
+
+Write-Host "=== AntiAtropos Local Teardown ===" -ForegroundColor Cyan
+
+# --- Stop port-forward ---
+Write-Host "[1/3] Stopping port-forwards..." -ForegroundColor Yellow
+$connections = Get-NetTCPConnection -LocalPort 3000 -ErrorAction SilentlyContinue
+if ($connections) {
+    $connections | ForEach-Object { Stop-Process -Id $_.OwningProcess -Force -ErrorAction SilentlyContinue }
+    Write-Host "  Stopped port-forward on :3000"
+}
+
+# --- Uninstall Helm releases ---
+Write-Host "[2/3] Uninstalling Helm releases..." -ForegroundColor Yellow
+helm uninstall grafana -n monitoring 2>&1 | Out-Null
+Write-Host "  Grafana uninstalled."
+helm uninstall prometheus -n monitoring 2>&1 | Out-Null
+Write-Host "  Prometheus uninstalled."
+
+# --- Delete namespaces ---
+Write-Host "[3/3] Deleting namespaces..." -ForegroundColor Yellow
+kubectl delete ns prod-sre monitoring 2>&1 | Out-Null
+
+Start-Sleep -Seconds 3
+Write-Host ""
+Write-Host "=== Teardown Complete ===" -ForegroundColor Green
+kubectl get ns