knowfoolery/backend/services/gateway-service/internal/middleware/health_checker.go

package middleware

import (
	"fmt"
	"net/http"
	"sync"
	"time"

	"github.com/gofiber/fiber/v3"
	
	"github.com/knowfoolery/backend/services/gateway-service/config"
)

type HealthChecker struct {
	services       map[string]*ServiceHealth
	config         config.ServicesConfig
	client         *http.Client
	mutex          sync.RWMutex
	lastCheck      time.Time
	checkInterval  time.Duration
}

type ServiceHealth struct {
	Name         string
	URL          string
	Status       ServiceStatus
	LastCheck    time.Time
	LastError    string
	ResponseTime time.Duration
	Uptime       time.Duration
	DownSince    time.Time
	CheckCount   int64
	FailCount    int64
}

type ServiceStatus string

const (
	ServiceStatusHealthy   ServiceStatus = "healthy"
	ServiceStatusUnhealthy ServiceStatus = "unhealthy"
	ServiceStatusUnknown   ServiceStatus = "unknown"
)

func NewHealthChecker(cfg config.ServicesConfig) *HealthChecker {
	services := make(map[string]*ServiceHealth)
	
	services["game-service"] = &ServiceHealth{
		Name:   "game-service",
		URL:    cfg.GameService.URL,
		Status: ServiceStatusUnknown,
	}
	services["question-service"] = &ServiceHealth{
		Name:   "question-service", 
		URL:    cfg.QuestionService.URL,
		Status: ServiceStatusUnknown,
	}
	services["user-service"] = &ServiceHealth{
		Name:   "user-service",
		URL:    cfg.UserService.URL, 
		Status: ServiceStatusUnknown,
	}
	services["leaderboard-service"] = &ServiceHealth{
		Name:   "leaderboard-service",
		URL:    cfg.LeaderboardService.URL,
		Status: ServiceStatusUnknown,
	}
	services["session-service"] = &ServiceHealth{
		Name:   "session-service",
		URL:    cfg.SessionService.URL,
		Status: ServiceStatusUnknown,
	}
	services["admin-service"] = &ServiceHealth{
		Name:   "admin-service",
		URL:    cfg.AdminService.URL,
		Status: ServiceStatusUnknown,
	}

	hc := &HealthChecker{
		services:      services,
		config:        cfg,
		client:        &http.Client{Timeout: 5 * time.Second},
		checkInterval: cfg.HealthCheckInterval,
	}

	go hc.startHealthChecks()
	
	return hc
}

func (hc *HealthChecker) Handler() fiber.Handler {
	return func(c fiber.Ctx) error {
		path := c.Path()
		
		serviceName := hc.getServiceNameFromPath(path)
		if serviceName == "" {
			return c.Next()
		}

		hc.mutex.RLock()
		service, exists := hc.services[serviceName]
		hc.mutex.RUnlock()

		if !exists {
			return c.Next()
		}

		if service.Status == ServiceStatusUnhealthy {
			return c.Status(fiber.StatusServiceUnavailable).JSON(fiber.Map{
				"error":   "Service temporarily unavailable",
				"service": serviceName,
				"status":  service.Status,
				"last_error": service.LastError,
			})
		}

		c.Set("X-Service-Status", string(service.Status))
		c.Set("X-Service-Response-Time", service.ResponseTime.String())
		
		return c.Next()
	}
}

func (hc *HealthChecker) startHealthChecks() {
	ticker := time.NewTicker(hc.checkInterval)
	defer ticker.Stop()

	for {
		select {
		case <-ticker.C:
			hc.checkAllServices()
		}
	}
}

func (hc *HealthChecker) checkAllServices() {
	var wg sync.WaitGroup
	
	hc.mutex.RLock()
	services := make([]*ServiceHealth, 0, len(hc.services))
	for _, service := range hc.services {
		services = append(services, service)
	}
	hc.mutex.RUnlock()

	for _, service := range services {
		wg.Add(1)
		go func(s *ServiceHealth) {
			defer wg.Done()
			hc.checkService(s)
		}(service)
	}

	wg.Wait()
	hc.lastCheck = time.Now()
}

func (hc *HealthChecker) checkService(service *ServiceHealth) {
	start := time.Now()
	
	healthURL := service.URL + "/health"
	
	resp, err := hc.client.Get(healthURL)
	responseTime := time.Since(start)
	
	hc.mutex.Lock()
	defer hc.mutex.Unlock()

	service.LastCheck = time.Now()
	service.ResponseTime = responseTime
	service.CheckCount++

	if err != nil {
		service.Status = ServiceStatusUnhealthy
		service.LastError = err.Error()
		service.FailCount++
		
		if service.DownSince.IsZero() {
			service.DownSince = time.Now()
		}
	} else {
		defer resp.Body.Close()
		
		if resp.StatusCode == http.StatusOK {
			previousStatus := service.Status
			service.Status = ServiceStatusHealthy
			service.LastError = ""
			
			if previousStatus == ServiceStatusUnhealthy && !service.DownSince.IsZero() {
				service.Uptime += time.Since(service.DownSince)
				service.DownSince = time.Time{}
			}
		} else {
			service.Status = ServiceStatusUnhealthy
			service.LastError = fmt.Sprintf("HTTP %d", resp.StatusCode)
			service.FailCount++
			
			if service.DownSince.IsZero() {
				service.DownSince = time.Now()
			}
		}
	}
}

func (hc *HealthChecker) getServiceNameFromPath(path string) string {
	if len(path) < 8 {
		return ""
	}
	
	if path[:8] != "/api/v1/" {
		return ""
	}

	pathParts := []string{}
	current := ""
	for i := 8; i < len(path); i++ {
		if path[i] == '/' {
			if current != "" {
				pathParts = append(pathParts, current)
				current = ""
			}
		} else {
			current += string(path[i])
		}
	}
	if current != "" {
		pathParts = append(pathParts, current)
	}

	if len(pathParts) == 0 {
		return ""
	}

	switch pathParts[0] {
	case "game":
		return "game-service"
	case "questions":
		return "question-service"
	case "users":
		return "user-service"
	case "leaderboard":
		return "leaderboard-service"
	case "sessions":
		return "session-service"
	case "admin":
		return "admin-service"
	default:
		return ""
	}
}

func (hc *HealthChecker) GetHealthStatus() map[string]interface{} {
	hc.mutex.RLock()
	defer hc.mutex.RUnlock()

	status := make(map[string]interface{})
	overallHealthy := true
	healthyCount := 0
	totalCount := len(hc.services)

	services := make(map[string]interface{})
	
	for name, service := range hc.services {
		serviceInfo := map[string]interface{}{
			"status":        service.Status,
			"url":          service.URL,
			"last_check":   service.LastCheck,
			"response_time": service.ResponseTime.String(),
			"check_count":  service.CheckCount,
			"fail_count":   service.FailCount,
		}

		if service.Status != ServiceStatusHealthy {
			overallHealthy = false
			serviceInfo["last_error"] = service.LastError
			if !service.DownSince.IsZero() {
				serviceInfo["down_since"] = service.DownSince
				serviceInfo["down_duration"] = time.Since(service.DownSince).String()
			}
		} else {
			healthyCount++
			if service.Uptime > 0 {
				serviceInfo["uptime"] = service.Uptime.String()
			}
		}

		if service.CheckCount > 0 {
			serviceInfo["success_rate"] = fmt.Sprintf("%.2f%%", 
				float64(service.CheckCount-service.FailCount)/float64(service.CheckCount)*100)
		}

		services[name] = serviceInfo
	}

	status["overall_status"] = "healthy"
	if !overallHealthy {
		status["overall_status"] = "degraded"
	}
	
	status["services"] = services
	status["summary"] = map[string]interface{}{
		"total_services":   totalCount,
		"healthy_services": healthyCount,
		"unhealthy_services": totalCount - healthyCount,
		"last_check":      hc.lastCheck,
	}

	return status
}

func (hc *HealthChecker) IsServiceHealthy(serviceName string) bool {
	hc.mutex.RLock()
	defer hc.mutex.RUnlock()

	if service, exists := hc.services[serviceName]; exists {
		return service.Status == ServiceStatusHealthy
	}
	
	return false
}

func (hc *HealthChecker) ForceHealthCheck() {
	go hc.checkAllServices()
}