feat: add embedded systemd installer and Prometheus metrics

Systemd Integration:
- New 'dbbackup install' command creates service/timer units
- Supports single-database and cluster backup modes
- Automatic dbbackup user/group creation with proper permissions
- Hardened service units with security features
- Template units with configurable OnCalendar schedules
- 'dbbackup uninstall' for clean removal

Prometheus Metrics:
- 'dbbackup metrics export' for textfile collector format
- 'dbbackup metrics serve' runs HTTP exporter on port 9399
- Metrics: last_success_timestamp, rpo_seconds, backup_total, etc.
- Integration with node_exporter textfile collector
- --with-metrics flag during install

Technical:
- Systemd templates embedded with //go:embed
- Service units include ReadWritePaths, OOMScoreAdjust
- Metrics exporter caches with 30s TTL
- Graceful shutdown on SIGTERM

cmd/install.go

@@ -0,0 +1,237 @@
+package cmd
+
+import (
+	"context"
+	"fmt"
+	"os"
+	"os/signal"
+	"strings"
+	"syscall"
+
+	"dbbackup/internal/installer"
+
+	"github.com/spf13/cobra"
+)
+
+var (
+	// Install flags
+	installInstance    string
+	installSchedule    string
+	installBackupType  string
+	installUser        string
+	installGroup       string
+	installBackupDir   string
+	installConfigPath  string
+	installTimeout     int
+	installWithMetrics bool
+	installMetricsPort int
+	installDryRun      bool
+	installStatus      bool
+
+	// Uninstall flags
+	uninstallPurge bool
+)
+
+// installCmd represents the install command
+var installCmd = &cobra.Command{
+	Use:   "install",
+	Short: "Install dbbackup as a systemd service",
+	Long: `Install dbbackup as a systemd service with automatic scheduling.
+
+This command creates systemd service and timer units for automated database backups.
+It supports both single database and cluster backup modes.
+
+Examples:
+  # Interactive installation (will prompt for options)
+  sudo dbbackup install
+
+  # Install cluster backup running daily at 2am
+  sudo dbbackup install --backup-type cluster --schedule "daily"
+
+  # Install single database backup with custom schedule
+  sudo dbbackup install --instance production --backup-type single --schedule "*-*-* 03:00:00"
+
+  # Install with Prometheus metrics exporter
+  sudo dbbackup install --with-metrics --metrics-port 9399
+
+  # Check installation status
+  dbbackup install --status
+
+  # Dry-run to see what would be installed
+  sudo dbbackup install --dry-run
+
+Schedule format (OnCalendar):
+  daily           - Every day at midnight
+  weekly          - Every Monday at midnight
+  *-*-* 02:00:00  - Every day at 2am
+  *-*-* 02,14:00  - Twice daily at 2am and 2pm
+  Mon *-*-* 03:00 - Every Monday at 3am
+`,
+	RunE: func(cmd *cobra.Command, args []string) error {
+		// Handle --status flag
+		if installStatus {
+			return runInstallStatus(cmd.Context())
+		}
+
+		return runInstall(cmd.Context())
+	},
+}
+
+// uninstallCmd represents the uninstall command
+var uninstallCmd = &cobra.Command{
+	Use:   "uninstall [instance]",
+	Short: "Uninstall dbbackup systemd service",
+	Long: `Uninstall dbbackup systemd service and timer.
+
+Examples:
+  # Uninstall default instance
+  sudo dbbackup uninstall
+
+  # Uninstall specific instance
+  sudo dbbackup uninstall production
+
+  # Uninstall and remove all configuration
+  sudo dbbackup uninstall --purge
+`,
+	RunE: func(cmd *cobra.Command, args []string) error {
+		instance := "cluster"
+		if len(args) > 0 {
+			instance = args[0]
+		}
+		return runUninstall(cmd.Context(), instance)
+	},
+}
+
+func init() {
+	rootCmd.AddCommand(installCmd)
+	rootCmd.AddCommand(uninstallCmd)
+
+	// Install flags
+	installCmd.Flags().StringVarP(&installInstance, "instance", "i", "", "Instance name (e.g., production, staging)")
+	installCmd.Flags().StringVarP(&installSchedule, "schedule", "s", "daily", "Backup schedule (OnCalendar format)")
+	installCmd.Flags().StringVarP(&installBackupType, "backup-type", "t", "cluster", "Backup type: single or cluster")
+	installCmd.Flags().StringVar(&installUser, "user", "dbbackup", "System user to run backups")
+	installCmd.Flags().StringVar(&installGroup, "group", "dbbackup", "System group for backup user")
+	installCmd.Flags().StringVar(&installBackupDir, "backup-dir", "/var/lib/dbbackup/backups", "Directory for backups")
+	installCmd.Flags().StringVar(&installConfigPath, "config-path", "/etc/dbbackup/dbbackup.conf", "Path to config file")
+	installCmd.Flags().IntVar(&installTimeout, "timeout", 3600, "Backup timeout in seconds")
+	installCmd.Flags().BoolVar(&installWithMetrics, "with-metrics", false, "Install Prometheus metrics exporter")
+	installCmd.Flags().IntVar(&installMetricsPort, "metrics-port", 9399, "Prometheus metrics port")
+	installCmd.Flags().BoolVar(&installDryRun, "dry-run", false, "Show what would be installed without making changes")
+	installCmd.Flags().BoolVar(&installStatus, "status", false, "Show installation status")
+
+	// Uninstall flags
+	uninstallCmd.Flags().BoolVar(&uninstallPurge, "purge", false, "Also remove configuration files")
+}
+
+func runInstall(ctx context.Context) error {
+	// Create context with signal handling
+	ctx, cancel := signal.NotifyContext(ctx, os.Interrupt, syscall.SIGTERM)
+	defer cancel()
+
+	// Expand schedule shortcuts
+	schedule := expandSchedule(installSchedule)
+
+	// Create installer
+	inst := installer.NewInstaller(log, installDryRun)
+
+	// Set up options
+	opts := installer.InstallOptions{
+		Instance:       installInstance,
+		BackupType:     installBackupType,
+		Schedule:       schedule,
+		User:           installUser,
+		Group:          installGroup,
+		BackupDir:      installBackupDir,
+		ConfigPath:     installConfigPath,
+		TimeoutSeconds: installTimeout,
+		WithMetrics:    installWithMetrics,
+		MetricsPort:    installMetricsPort,
+	}
+
+	// For cluster backup, override instance
+	if installBackupType == "cluster" {
+		opts.Instance = "cluster"
+	}
+
+	return inst.Install(ctx, opts)
+}
+
+func runUninstall(ctx context.Context, instance string) error {
+	ctx, cancel := signal.NotifyContext(ctx, os.Interrupt, syscall.SIGTERM)
+	defer cancel()
+
+	inst := installer.NewInstaller(log, false)
+	return inst.Uninstall(ctx, instance, uninstallPurge)
+}
+
+func runInstallStatus(ctx context.Context) error {
+	inst := installer.NewInstaller(log, false)
+
+	// Check cluster status
+	clusterStatus, err := inst.Status(ctx, "cluster")
+	if err != nil {
+		return err
+	}
+
+	fmt.Println()
+	fmt.Println("📦 DBBackup Installation Status")
+	fmt.Println(strings.Repeat("═", 50))
+
+	if clusterStatus.Installed {
+		fmt.Println()
+		fmt.Println("🔹 Cluster Backup:")
+		fmt.Printf("   Service:  %s\n", formatStatus(clusterStatus.Installed, clusterStatus.Active))
+		fmt.Printf("   Timer:    %s\n", formatStatus(clusterStatus.TimerEnabled, clusterStatus.TimerActive))
+		if clusterStatus.NextRun != "" {
+			fmt.Printf("   Next run: %s\n", clusterStatus.NextRun)
+		}
+		if clusterStatus.LastRun != "" {
+			fmt.Printf("   Last run: %s\n", clusterStatus.LastRun)
+		}
+	} else {
+		fmt.Println()
+		fmt.Println("❌ No systemd services installed")
+		fmt.Println()
+		fmt.Println("Run 'sudo dbbackup install' to install as a systemd service")
+	}
+
+	// Check for exporter
+	if _, err := os.Stat("/etc/systemd/system/dbbackup-exporter.service"); err == nil {
+		exporterStatus, err := inst.Status(ctx, "exporter")
+		fmt.Println()
+		fmt.Println("🔹 Metrics Exporter:")
+		if err == nil && exporterStatus != nil {
+			fmt.Printf("   Service:  %s\n", formatStatus(true, exporterStatus.Active))
+		} else {
+			fmt.Printf("   Service:  installed (status unknown)\n")
+		}
+	}
+
+	fmt.Println()
+	return nil
+}
+
+func formatStatus(installed, active bool) string {
+	if !installed {
+		return "not installed"
+	}
+	if active {
+		return "✅ active"
+	}
+	return "⚪ inactive"
+}
+
+func expandSchedule(schedule string) string {
+	shortcuts := map[string]string{
+		"hourly":  "*-*-* *:00:00",
+		"daily":   "*-*-* 02:00:00",
+		"weekly":  "Mon *-*-* 02:00:00",
+		"monthly": "*-*-01 02:00:00",
+	}
+
+	if expanded, ok := shortcuts[strings.ToLower(schedule)]; ok {
+		return expanded
+	}
+	return schedule
+}

cmd/metrics.go

@@ -0,0 +1,138 @@
+package cmd
+
+import (
+	"context"
+	"fmt"
+	"os"
+	"os/signal"
+	"syscall"
+
+	"dbbackup/internal/prometheus"
+
+	"github.com/spf13/cobra"
+)
+
+var (
+	metricsInstance string
+	metricsOutput   string
+	metricsPort     int
+)
+
+// metricsCmd represents the metrics command
+var metricsCmd = &cobra.Command{
+	Use:   "metrics",
+	Short: "Prometheus metrics management",
+	Long: `Prometheus metrics management for dbbackup.
+
+Export metrics to a textfile for node_exporter, or run an HTTP server
+for direct Prometheus scraping.`,
+}
+
+// metricsExportCmd exports metrics to a textfile
+var metricsExportCmd = &cobra.Command{
+	Use:   "export",
+	Short: "Export metrics to textfile",
+	Long: `Export Prometheus metrics to a textfile for node_exporter.
+
+The textfile collector in node_exporter can scrape metrics from files
+in a designated directory (typically /var/lib/node_exporter/textfile_collector/).
+
+Examples:
+  # Export metrics to default location
+  dbbackup metrics export
+
+  # Export with custom output path
+  dbbackup metrics export --output /var/lib/dbbackup/metrics/dbbackup.prom
+
+  # Export for specific instance
+  dbbackup metrics export --instance production --output /var/lib/dbbackup/metrics/production.prom
+
+After export, configure node_exporter with:
+  --collector.textfile.directory=/var/lib/dbbackup/metrics/
+`,
+	RunE: func(cmd *cobra.Command, args []string) error {
+		return runMetricsExport(cmd.Context())
+	},
+}
+
+// metricsServeCmd runs the HTTP metrics server
+var metricsServeCmd = &cobra.Command{
+	Use:   "serve",
+	Short: "Run Prometheus HTTP server",
+	Long: `Run an HTTP server exposing Prometheus metrics.
+
+This starts a long-running daemon that serves metrics at /metrics.
+Prometheus can scrape this endpoint directly.
+
+Examples:
+  # Start server on default port 9399
+  dbbackup metrics serve
+
+  # Start server on custom port
+  dbbackup metrics serve --port 9100
+
+  # Run as systemd service (installed via 'dbbackup install --with-metrics')
+  sudo systemctl start dbbackup-exporter
+
+Endpoints:
+  /metrics  - Prometheus metrics
+  /health   - Health check (returns 200 OK)
+  /         - Service info page
+`,
+	RunE: func(cmd *cobra.Command, args []string) error {
+		return runMetricsServe(cmd.Context())
+	},
+}
+
+func init() {
+	rootCmd.AddCommand(metricsCmd)
+	metricsCmd.AddCommand(metricsExportCmd)
+	metricsCmd.AddCommand(metricsServeCmd)
+
+	// Export flags
+	metricsExportCmd.Flags().StringVar(&metricsInstance, "instance", "default", "Instance name for metrics labels")
+	metricsExportCmd.Flags().StringVarP(&metricsOutput, "output", "o", "/var/lib/dbbackup/metrics/dbbackup.prom", "Output file path")
+
+	// Serve flags
+	metricsServeCmd.Flags().StringVar(&metricsInstance, "instance", "default", "Instance name for metrics labels")
+	metricsServeCmd.Flags().IntVarP(&metricsPort, "port", "p", 9399, "HTTP server port")
+}
+
+func runMetricsExport(ctx context.Context) error {
+	// Open catalog
+	cat, err := openCatalog()
+	if err != nil {
+		return fmt.Errorf("failed to open catalog: %w", err)
+	}
+	defer cat.Close()
+
+	// Create metrics writer
+	writer := prometheus.NewMetricsWriter(log, cat, metricsInstance)
+
+	// Write textfile
+	if err := writer.WriteTextfile(metricsOutput); err != nil {
+		return fmt.Errorf("failed to write metrics: %w", err)
+	}
+
+	log.Info("Exported metrics to textfile", "path", metricsOutput, "instance", metricsInstance)
+	return nil
+}
+
+func runMetricsServe(ctx context.Context) error {
+	// Setup signal handling
+	ctx, cancel := signal.NotifyContext(ctx, os.Interrupt, syscall.SIGTERM)
+	defer cancel()
+
+	// Open catalog
+	cat, err := openCatalog()
+	if err != nil {
+		return fmt.Errorf("failed to open catalog: %w", err)
+	}
+	defer cat.Close()
+
+	// Create exporter
+	exporter := prometheus.NewExporter(log, cat, metricsInstance, metricsPort)
+
+	// Run server (blocks until context is cancelled)
+	return exporter.Serve(ctx)
+}