Merge pull request #382 from overmindtech/add-cpu-load-simulation

feat: add 70% CPU load simulation for realistic demo metrics

Author: jameslaneovermind

modules/api-server/README.md

@@ -41,6 +41,16 @@ Overmind analyzes the PR and catches the hidden risk.
 
 **The trap**: Both instances look identical on paper, but t3 uses CPU credits. At 70% sustained CPU, credits exhaust and performance collapses.
 
+## 🔥 CPU Load Simulation
+
+The instance runs a **permanent 70% CPU load** via `stress-ng` to simulate a real CPU-intensive workload. This ensures:
+
+- CloudWatch shows realistic CPU utilization (~70%)
+- Overmind can detect the risk from actual metrics
+- The demo reflects real-world conditions
+
+The load runs as a systemd service that starts on boot and restarts automatically. No extra cost - EC2 is billed by time, not CPU usage.
+
 ## 🔍 What Overmind Catches
 
 1. **Blast Radius** - ALB, target groups, alarms, security groups

modules/api-server/ec2.tf

@@ -25,7 +25,28 @@ resource "aws_instance" "api_server" {
   user_data = base64encode(<<-EOF
     #!/bin/bash
     yum update -y
-    yum install -y httpd
+    yum install -y httpd stress-ng
+
+    # Create systemd service for CPU load simulation (70% sustained)
+    cat > /etc/systemd/system/cpu-load-simulator.service <<'SERVICE'
+    [Unit]
+    Description=Simulate 70% CPU load for demo purposes
+    After=network.target
+
+    [Service]
+    Type=simple
+    ExecStart=/usr/bin/stress-ng --cpu 2 --cpu-load 70
+    Restart=always
+    RestartSec=10
+
+    [Install]
+    WantedBy=multi-user.target
+    SERVICE
+
+    # Enable and start CPU load simulation
+    systemctl daemon-reload
+    systemctl enable cpu-load-simulator
+    systemctl start cpu-load-simulator
 
     cat > /var/www/html/health <<'HEALTH'
     OK