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>

scripts/docker-setup.sh

@@ -1512,6 +1512,16 @@ print_summary() {
         echo -e "${YELLOW}Env vars (passwords, secrets) must be filled in manually.${NC}"
     fi
     echo ""
+    if [[ -n "$CUSTOMER_ID" ]]; then
+        echo -e "${BOLD}${YELLOW}Disaster Recovery:${NC}"
+        echo "  If this is a reinstallation, the controller will automatically:"
+        echo "    1. Contact the Hub for your previous configuration"
+        echo "    2. Mount your existing storage drives"
+        echo "    3. Detect and offer to restore your applications"
+        echo ""
+        echo "  Open https://felhom.${BASE_DOMAIN} to monitor the restore process."
+        echo ""
+    fi
     echo -e "${BOLD}Quick Commands:${NC}"
     echo "  dps          → List containers"
     echo "  dlogs <n>    → View container logs"