v0.15.5: Disaster recovery — Hub-based infra backup, auto-mount, restore UI

Complete DR implementation (TASK2.md Phases 1-4): - Hub infra-backup push/pull endpoints (controller.yaml, disk layout, stacks) - Fresh-deployment detection pulls config from Hub, auto-mounts drives by UUID - Full-page restore UI with drive status, app table, sequential restore - docker-setup.sh shows DR instructions when customer_id is configured New files: disk_layout.go, restore_scan.go, restore_app_linux.go, restore_drives_linux.go, infra_backup.go, infra_pull.go, handler_restore.go, restore.html Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-19 13:16:46 +01:00
parent 5d993b66a2
commit 6713df2186
21 changed files with 3324 additions and 9 deletions
@@ -0,0 +1,132 @@
+package web
+
+import (
+	"context"
+	"encoding/json"
+	"net/http"
+	"time"
+
+	"gitea.dooplex.hu/admin/felhom-controller/internal/backup"
+)
+
+// restorePageHandler renders the full-page DR restore UI.
+func (s *Server) restorePageHandler(w http.ResponseWriter, r *http.Request) {
+	if s.restorePlan == nil {
+		http.Redirect(w, r, "/", http.StatusFound)
+		return
+	}
+
+	data := map[string]interface{}{
+		"Title":        "Katasztrófa utáni visszaállítás",
+		"CustomerName": s.cfg.Customer.Name,
+		"Domain":       s.cfg.Customer.Domain,
+		"Version":      s.version,
+		"CustomerID":   s.restorePlan.CustomerID,
+		"Timestamp":    s.restorePlan.Timestamp,
+		"Apps":         s.restorePlan.GetApps(),
+		"Drives":       s.restorePlan.Drives,
+		"PlanStatus":   s.restorePlan.Status,
+	}
+
+	s.render(w, "restore", data)
+}
+
+// apiRestoreStatus returns the current restore plan status as JSON.
+func (s *Server) apiRestoreStatus(w http.ResponseWriter, r *http.Request) {
+	if s.restorePlan == nil {
+		jsonError(w, "not in restore mode", http.StatusBadRequest)
+		return
+	}
+
+	w.Header().Set("Content-Type", "application/json; charset=utf-8")
+	json.NewEncoder(w).Encode(s.restorePlan.Snapshot())
+}
+
+// apiRestoreAll starts restoring all pending apps sequentially.
+func (s *Server) apiRestoreAll(w http.ResponseWriter, r *http.Request) {
+	if s.restorePlan == nil {
+		jsonError(w, "not in restore mode", http.StatusBadRequest)
+		return
+	}
+	if s.restorePlan.Status == "restoring" {
+		jsonError(w, "restore already in progress", http.StatusConflict)
+		return
+	}
+
+	s.restorePlan.Status = "restoring"
+	go s.executeAllRestores()
+
+	jsonResponse(w, map[string]interface{}{
+		"ok":      true,
+		"message": "Visszaállítás elindítva",
+	})
+}
+
+// apiRestoreSkip exits restore mode without restoring.
+func (s *Server) apiRestoreSkip(w http.ResponseWriter, r *http.Request) {
+	if s.restorePlan == nil {
+		jsonError(w, "not in restore mode", http.StatusBadRequest)
+		return
+	}
+
+	s.logger.Println("[INFO] User skipped DR restore — entering normal mode")
+	s.clearRestoreMode()
+
+	jsonResponse(w, map[string]interface{}{
+		"ok":      true,
+		"message": "Visszaállítás kihagyva",
+	})
+}
+
+// executeAllRestores runs the restore for each pending app sequentially.
+func (s *Server) executeAllRestores() {
+	s.logger.Println("[INFO] Starting DR restore for all apps")
+
+	for i := range s.restorePlan.Apps {
+		app := &s.restorePlan.Apps[i]
+		if app.Status != "pending" {
+			continue
+		}
+
+		s.restorePlan.UpdateApp(app.Name, "restoring", "")
+		s.logger.Printf("[INFO] Restoring app %s (%s)", app.Name, app.DisplayName)
+
+		ctx, cancel := context.WithTimeout(context.Background(), 10*time.Minute)
+		err := backup.RestoreAppFromBackup(ctx, app, s.cfg.Paths.StacksDir, s.logger)
+		cancel()
+
+		if err != nil {
+			s.restorePlan.UpdateApp(app.Name, "failed", err.Error())
+			s.logger.Printf("[ERROR] Restore failed for %s: %v", app.Name, err)
+		} else {
+			s.restorePlan.UpdateApp(app.Name, "done", "")
+			s.logger.Printf("[INFO] Restore completed for %s", app.Name)
+		}
+	}
+
+	s.restorePlan.Status = "done"
+	s.logger.Println("[INFO] All app restores completed")
+
+	// Re-scan stacks so dashboard picks up restored apps
+	if s.stackMgr != nil {
+		if err := s.stackMgr.ScanStacks(); err != nil {
+			s.logger.Printf("[WARN] Post-restore stack scan failed: %v", err)
+		}
+	}
+
+	// Auto-clear restore mode after a brief delay so user can see final status
+	go func() {
+		time.Sleep(5 * time.Second)
+		// Only auto-clear if user hasn't already navigated away
+		if s.restorePlan != nil && s.restorePlan.AllDone() {
+			// Keep plan visible — user clicks "continue to dashboard" to clear
+		}
+	}()
+}
+
+// clearRestoreMode exits restore mode and returns to normal operation.
+func (s *Server) clearRestoreMode() {
+	s.restoreMu.Lock()
+	defer s.restoreMu.Unlock()
+	s.restorePlan = nil
+}