edgeguard-native/internal/handlers/ntp.go

package handlers

import (
	"context"
	"errors"
	"fmt"
	"log/slog"
	"os/exec"
	"strings"

	"github.com/gin-gonic/gin"

	"git.netcell-it.de/projekte/edgeguard-native/internal/handlers/response"
	"git.netcell-it.de/projekte/edgeguard-native/internal/models"
	"git.netcell-it.de/projekte/edgeguard-native/internal/services/audit"
	ntpsvc "git.netcell-it.de/projekte/edgeguard-native/internal/services/ntp"
)

type NTPHandler struct {
	Repo     *ntpsvc.Repo
	Audit    *audit.Repo
	NodeID   string
	Reloader func(ctx context.Context) error
}

func NewNTPHandler(repo *ntpsvc.Repo, a *audit.Repo, nodeID string, reloader func(context.Context) error) *NTPHandler {
	return &NTPHandler{Repo: repo, Audit: a, NodeID: nodeID, Reloader: reloader}
}

func (h *NTPHandler) reload(ctx context.Context, op string) {
	if h.Reloader == nil {
		return
	}
	if err := h.Reloader(ctx); err != nil {
		slog.Warn("chrony: reload after mutation failed", "op", op, "error", err)
	}
}

func (h *NTPHandler) Register(rg *gin.RouterGroup) {
	g := rg.Group("/ntp")
	g.GET("/settings", h.GetSettings)
	g.PUT("/settings", h.UpdateSettings)
	g.GET("/status", h.Status)
	g.POST("/force-sync", h.ForceSync)

	p := g.Group("/pools")
	p.GET("", h.ListPools)
	p.POST("", h.CreatePool)
	p.GET("/:id", h.GetPool)
	p.PUT("/:id", h.UpdatePool)
	p.DELETE("/:id", h.DeletePool)
}

// Status liefert den aktuellen chrony-Tracking-Status via `chronyc tracking`.
// Output-Felder: synced (bool), reference (Quell-Server), stratum (int),
// offset_ms (float), freq_ppm (float), rms_offset_ms (float), error (string
// wenn chronyc nicht verfügbar oder System-Takt nicht sync'd).
func (h *NTPHandler) Status(c *gin.Context) {
	out, err := exec.Command("chronyc", "tracking").Output()
	if err != nil {
		response.OK(c, gin.H{
			"synced":    false,
			"error":     "chronyc nicht verfügbar: " + err.Error(),
		})
		return
	}
	resp := parseChronyTracking(string(out))
	response.OK(c, resp)
}

type chronyStatus struct {
	Synced       bool    `json:"synced"`
	Reference    string  `json:"reference"`
	Stratum      int     `json:"stratum"`
	OffsetMs     float64 `json:"offset_ms"`
	FreqPPM      float64 `json:"freq_ppm"`
	RMSOffsetMs  float64 `json:"rms_offset_ms"`
	Error        string  `json:"error,omitempty"`
}

func parseChronyTracking(out string) chronyStatus {
	s := chronyStatus{}
	for _, line := range strings.Split(out, "\n") {
		line = strings.TrimSpace(line)
		key, val, ok := strings.Cut(line, ":")
		if !ok {
			continue
		}
		key = strings.TrimSpace(key)
		val = strings.TrimSpace(val)
		switch key {
		case "Reference ID":
			// "Reference ID    : A29FC801 (time.cloudflare.com)"
			if i := strings.Index(val, "("); i >= 0 {
				s.Reference = strings.Trim(val[i:], "()")
			} else {
				s.Reference = val
			}
			s.Synced = val != "00000000 ()"
		case "Stratum":
			fmt.Sscanf(val, "%d", &s.Stratum)
			if s.Stratum > 0 && s.Stratum < 16 {
				s.Synced = true
			}
		case "System time":
			// "0.000012345 seconds fast of NTP time"
			var v float64
			fmt.Sscanf(val, "%f", &v)
			s.OffsetMs = v * 1000
		case "Frequency":
			// "-12.345 ppm slow" or "+12.345 ppm fast"
			fmt.Sscanf(val, "%f", &s.FreqPPM)
		case "RMS offset":
			var v float64
			fmt.Sscanf(val, "%f", &v)
			s.RMSOffsetMs = v * 1000
		}
	}
	return s
}

func (h *NTPHandler) GetSettings(c *gin.Context) {
	s, err := h.Repo.GetSettings(c.Request.Context())
	if err != nil {
		response.Internal(c, err)
		return
	}
	response.OK(c, s)
}

func (h *NTPHandler) UpdateSettings(c *gin.Context) {
	var req models.NTPSettings
	if err := c.ShouldBindJSON(&req); err != nil {
		response.BadRequest(c, err)
		return
	}
	out, err := h.Repo.UpdateSettings(c.Request.Context(), req)
	if err != nil {
		response.Internal(c, err)
		return
	}
	_ = h.Audit.Log(c.Request.Context(), actorOf(c), "ntp.settings.update", "settings", out, h.NodeID)
	response.OK(c, out)
	h.reload(c.Request.Context(), "settings.update")
}

func (h *NTPHandler) ListPools(c *gin.Context) {
	out, err := h.Repo.ListPools(c.Request.Context())
	if err != nil {
		response.Internal(c, err)
		return
	}
	response.OK(c, gin.H{"pools": out})
}

func (h *NTPHandler) GetPool(c *gin.Context) {
	id, ok := parseID(c)
	if !ok {
		return
	}
	p, err := h.Repo.GetPool(c.Request.Context(), id)
	if err != nil {
		if errors.Is(err, ntpsvc.ErrPoolNotFound) {
			response.NotFound(c, err)
			return
		}
		response.Internal(c, err)
		return
	}
	response.OK(c, p)
}

func (h *NTPHandler) CreatePool(c *gin.Context) {
	var req models.NTPPool
	if err := c.ShouldBindJSON(&req); err != nil {
		response.BadRequest(c, err)
		return
	}
	if err := validateNTPPool(&req); err != nil {
		response.BadRequest(c, err)
		return
	}
	out, err := h.Repo.CreatePool(c.Request.Context(), req)
	if err != nil {
		response.Internal(c, err)
		return
	}
	_ = h.Audit.Log(c.Request.Context(), actorOf(c), "ntp.pool.create", out.Address, out, h.NodeID)
	response.Created(c, out)
	h.reload(c.Request.Context(), "pool.create")
}

func (h *NTPHandler) UpdatePool(c *gin.Context) {
	id, ok := parseID(c)
	if !ok {
		return
	}
	var req models.NTPPool
	if err := c.ShouldBindJSON(&req); err != nil {
		response.BadRequest(c, err)
		return
	}
	if err := validateNTPPool(&req); err != nil {
		response.BadRequest(c, err)
		return
	}
	out, err := h.Repo.UpdatePool(c.Request.Context(), id, req)
	if err != nil {
		if errors.Is(err, ntpsvc.ErrPoolNotFound) {
			response.NotFound(c, err)
			return
		}
		response.Internal(c, err)
		return
	}
	_ = h.Audit.Log(c.Request.Context(), actorOf(c), "ntp.pool.update", out.Address, out, h.NodeID)
	response.OK(c, out)
	h.reload(c.Request.Context(), "pool.update")
}

func (h *NTPHandler) DeletePool(c *gin.Context) {
	id, ok := parseID(c)
	if !ok {
		return
	}
	if err := h.Repo.DeletePool(c.Request.Context(), id); err != nil {
		if errors.Is(err, ntpsvc.ErrPoolNotFound) {
			response.NotFound(c, err)
			return
		}
		response.Internal(c, err)
		return
	}
	response.NoContent(c)
	h.reload(c.Request.Context(), "pool.delete")
}

// ForceSync runs `chronyc makestep` which immediately adjusts the
// system clock to the current NTP reference. Useful after a long
// outage or VM migration where the clock has drifted by more than
// the 1ms default slew threshold.
func (h *NTPHandler) ForceSync(c *gin.Context) {
	out, err := exec.Command("chronyc", "makestep").CombinedOutput()
	if err != nil {
		slog.Error("ntp force-sync failed", "err", err, "out", string(out))
		response.Internal(c, err)
		return
	}
	_ = h.Audit.Log(c.Request.Context(), actorOf(c), "ntp.force-sync", "chrony", nil, h.NodeID)
	response.OK(c, gin.H{"message": "clock stepped", "output": string(out)})
}

func validateNTPPool(p *models.NTPPool) error {
	if p.Address == "" {
		return errors.New("address required")
	}
	switch p.Kind {
	case "pool", "server":
	default:
		return errors.New("kind must be 'pool' or 'server'")
	}
	return nil
}