Proxmox MCP Server

#!/bin/bash # Health Monitoring Script for Proxmox MCP Server # This script monitors the deployment and provides health status set -euo pipefail # Configuration DEPLOY_DIR="/opt/proxmox-mcp" SERVICE_NAME="proxmox-mcp" LOG_FILE="/var/log/proxmox-mcp-monitor.log" # Colors for output RED='\033[0;31m' GREEN='\033[0;32m' YELLOW='\033[1;33m' BLUE='\033[0;34m' NC='\033[0m' log() { local message="[$(date +'%Y-%m-%d %H:%M:%S')] $1" echo -e "${GREEN}$message${NC}" echo "$message" >> "$LOG_FILE" 2>/dev/null || true } warn() { local message="[WARNING] $1" echo -e "${YELLOW}$message${NC}" echo "$message" >> "$LOG_FILE" 2>/dev/null || true } error() { local message="[ERROR] $1" echo -e "${RED}$message${NC}" echo "$message" >> "$LOG_FILE" 2>/dev/null || true } info() { local message="[INFO] $1" echo -e "${BLUE}$message${NC}" echo "$message" >> "$LOG_FILE" 2>/dev/null || true } # Check if containers are running check_containers() { local status="healthy" local details="" if [[ ! -d "$DEPLOY_DIR" ]]; then error "❌ Deployment directory not found: $DEPLOY_DIR" return 1 fi cd "$DEPLOY_DIR" # Check if docker-compose.yml exists if [[ ! -f "docker-compose.yml" ]]; then error "❌ docker-compose.yml not found in $DEPLOY_DIR" return 1 fi # Get container status local containers=$(docker-compose ps -q 2>/dev/null || echo "") if [[ -z "$containers" ]]; then error "❌ No containers found" return 1 fi # Check each container while read -r container_id; do if [[ -n "$container_id" ]]; then local container_name=$(docker inspect --format='{{.Name}}' "$container_id" 2>/dev/null | sed 's/^.//') local container_status=$(docker inspect --format='{{.State.Status}}' "$container_id" 2>/dev/null) local container_health=$(docker inspect --format='{{.State.Health.Status}}' "$container_id" 2>/dev/null || echo "none") if [[ "$container_status" == "running" ]]; then if [[ "$container_health" == "healthy" ]] || [[ "$container_health" == "none" ]]; then log "✅ $container_name: running" else warn "⚠️ $container_name: running but unhealthy ($container_health)" status="degraded" fi else error "❌ $container_name: $container_status" status="unhealthy" fi fi done <<< "$containers" if [[ "$status" == "healthy" ]]; then log "✅ All containers are healthy" return 0 elif [[ "$status" == "degraded" ]]; then warn "⚠️ Some containers are degraded" return 1 else error "❌ Containers are unhealthy" return 2 fi } # Check service endpoints check_endpoints() { local endpoints_ok=true # Check reverse proxy (port 80) if curl -f -s --max-time 10 http://localhost/docs > /dev/null; then log "✅ Reverse proxy (port 80): accessible" else error "❌ Reverse proxy (port 80): not accessible" endpoints_ok=false fi # Check direct MCP server (port 8080) if curl -f -s --max-time 10 http://localhost:8080/docs > /dev/null; then log "✅ Direct MCP server (port 8080): accessible" else error "❌ Direct MCP server (port 8080): not accessible" endpoints_ok=false fi # Check health endpoint (may return unhealthy but should respond) local health_response=$(curl -s --max-time 10 http://localhost/health 2>/dev/null || echo "no response") if [[ "$health_response" != "no response" ]]; then log "✅ Health endpoint: responding" info "Health status: $health_response" else error "❌ Health endpoint: not responding" endpoints_ok=false fi if $endpoints_ok; then log "✅ All endpoints are accessible" return 0 else error "❌ Some endpoints are not accessible" return 1 fi } # Check resource usage check_resources() { if [[ ! -d "$DEPLOY_DIR" ]]; then warn "Cannot check resources - deployment directory not found" return 1 fi cd "$DEPLOY_DIR" info "Resource usage:" # Get container resource usage local containers=$(docker-compose ps -q 2>/dev/null || echo "") if [[ -n "$containers" ]]; then while read -r container_id; do if [[ -n "$container_id" ]]; then local name=$(docker inspect --format='{{.Name}}' "$container_id" 2>/dev/null | sed 's/^.//') local stats=$(docker stats --no-stream --format "table {{.CPUPerc}}\t{{.MemUsage}}" "$container_id" 2>/dev/null | tail -n 1) info " $name: $stats" fi done <<< "$containers" fi # System resources local cpu_usage=$(top -bn1 | grep "Cpu(s)" | awk '{print $2}' | cut -d'%' -f1) local mem_usage=$(free | grep Mem | awk '{printf "%.1f", $3/$2 * 100.0}') local disk_usage=$(df -h "$DEPLOY_DIR" | awk 'NR==2{print $5}') info " System CPU: ${cpu_usage}%" info " System Memory: ${mem_usage}%" info " Disk usage: $disk_usage" } # Show service logs show_logs() { if [[ ! -d "$DEPLOY_DIR" ]]; then error "Deployment directory not found: $DEPLOY_DIR" return 1 fi cd "$DEPLOY_DIR" local lines="${1:-50}" info "Showing last $lines lines of logs:" docker-compose logs --tail="$lines" || error "Failed to get logs" } # Restart services restart_services() { if [[ ! -d "$DEPLOY_DIR" ]]; then error "Deployment directory not found: $DEPLOY_DIR" return 1 fi info "Restarting services..." cd "$DEPLOY_DIR" docker-compose restart || error "Failed to restart services" # Wait a moment for services to start sleep 5 log "✅ Services restarted" } # Auto-heal function auto_heal() { log "🔧 Starting auto-heal process..." # First try a simple restart if restart_services; then sleep 10 if check_containers && check_endpoints; then log "✅ Auto-heal successful - services recovered" return 0 fi fi # If restart didn't work, try recreating containers warn "Restart didn't work, trying to recreate containers..." cd "$DEPLOY_DIR" docker-compose down --timeout 30 docker-compose up -d --remove-orphans sleep 15 if check_containers && check_endpoints; then log "✅ Auto-heal successful - containers recreated" return 0 else error "❌ Auto-heal failed - manual intervention required" return 1 fi } # Full health check full_health_check() { log "🔍 Starting comprehensive health check..." local overall_status="healthy" # Check containers if ! check_containers; then overall_status="unhealthy" fi # Check endpoints if ! check_endpoints; then overall_status="unhealthy" fi # Show resources check_resources echo "" if [[ "$overall_status" == "healthy" ]]; then log "🎉 Overall status: HEALTHY" return 0 else error "💥 Overall status: UNHEALTHY" return 1 fi } # Watch mode - continuous monitoring watch_mode() { local interval="${1:-60}" log "👀 Starting watch mode (checking every ${interval}s)..." log "Press Ctrl+C to stop" while true; do echo "$(date +'%Y-%m-%d %H:%M:%S') - Health Check" echo "==================================" if full_health_check; then echo "✅ Status: OK" else echo "❌ Status: ISSUES DETECTED" warn "Issues detected - consider running auto-heal" fi echo "" sleep "$interval" done } # Main function main() { case "${1:-check}" in "check"|"status") full_health_check ;; "containers") check_containers ;; "endpoints") check_endpoints ;; "resources") check_resources ;; "logs") show_logs "${2:-50}" ;; "restart") restart_services ;; "heal") auto_heal ;; "watch") watch_mode "${2:-60}" ;; *) echo "Usage: $0 [command] [options]" echo "" echo "Commands:" echo " check - Full health check (default)" echo " status - Same as check" echo " containers - Check container status only" echo " endpoints - Check service endpoints only" echo " resources - Show resource usage" echo " logs [n] - Show last n lines of logs (default: 50)" echo " restart - Restart all services" echo " heal - Auto-heal services (restart/recreate)" echo " watch [s] - Continuous monitoring (default: 60s interval)" echo "" echo "Examples:" echo " $0 # Quick health check" echo " $0 logs 100 # Show last 100 log lines" echo " $0 watch 30 # Monitor every 30 seconds" exit 1 ;; esac } # Ensure log file directory exists mkdir -p "$(dirname "$LOG_FILE")" 2>/dev/null || true # Run main function main "$@"