1.0 Problem Summary

This document provides a comprehensive technical analysis and resolution path for a specific Disaster Recovery System (DRS) backup failure on the Cisco IM and Presence (IM&P) platform, specifically observed on version 14.x.

The core symptom is the failure of a scheduled or manual backup job. Analysis of the DRF Master logs (drf_master.log) on the IM&P Publisher node reveals the following critical error message during the backup estimation phase:

file:/common/drf/localDevice/drfComponent.xml, unable to read file.

This error indicates that the DRF Master Agent, which orchestrates the backup process across the cluster, is unable to access a key configuration file required to communicate with local backup components. The failure prevents the backup from proceeding beyond the initial estimation stage.

2.0 Root Cause Analysis

The drfComponent.xml file, located in the /common/drf/localDevice/ directory, serves as a dynamically generated manifest. The DRF Master Agent uses this file to identify and query all registered components on the local node (e.g., IM&P database, TFTP files) to estimate the total size and scope of the required backup.

The “unable to read file” error typically stems from one of the following conditions:

• File Corruption: The drfComponent.xml file has become corrupted due to a disk error, an improper shutdown, or a software anomaly.
• File Permissions: Incorrect file system permissions may prevent the DRF Master process from accessing the file.
• Missing File: The file may have been inadvertently deleted or failed to generate correctly after a service restart or system upgrade.

Because this file is essential for the initial handshake between the master and local backup agents, any issue with its integrity or accessibility will cause an immediate failure of the backup process.

3.0 Pre-Resolution Diagnostics

Before proceeding with the resolution, it is imperative to perform the following diagnostic steps to confirm the scope of the issue and rule out other common backup-related problems.

1. Verify SFTP Server Connectivity: From the CUCM Publisher’s DRS interface (which manages the backup schedule for the entire cluster), verify the SFTP server configuration. Use the “Test” button to ensure there are no underlying network, credential, or path permission issues with the backup target.
2. Check Service Status: Log in to the Cisco Unified Serviceability page for both the IM&P Publisher and Subscriber nodes. Navigate to Tools > Control Center - Feature Services. Ensure the Cisco DRF Master Agent (on the Publisher) and the Cisco DRF Local Agent (on all nodes) services are running and activated.
3. Inspect the Problematic File via CLI:
   • Establish an SSH session to the IM&P Publisher node.
   • Execute the command file list activelog /common/drf/localDevice/ to check for the existence of drfComponent.xml.
   • If the file exists, attempt to view its contents with file view activelog /common/drf/localDevice/drfComponent.xml. If this command returns garbled text or an I/O error, it strongly suggests file corruption.

4.0 Comprehensive Resolution Procedure

This procedure details the steps to force the regeneration of the drfComponent.xml file. This should be performed during a scheduled maintenance window, as it involves restarting a core system service.

1. Establish CLI Access: Open an SSH session to the IM&P Publisher node using administrative credentials.

2. Delete the Corrupt Manifest File: Execute the following command precisely. This action removes the problematic file, which will trigger its recreation upon service restart.

   file delete activelog /common/drf/localDevice/drfComponent.xml
   

   Confirm the deletion when prompted.

3. Restart the DRF Master Agent Service: The recommended method is via the CLI to ensure a clean restart.

   utils service restart Cisco DRF Master Agent
   

   This service is responsible for regenerating the drfComponent.xml file upon startup. The restart process may take several minutes.

4. Verify File Regeneration: After the service restart command completes, wait approximately 2-3 minutes. Then, execute the following command to confirm the file has been successfully recreated:

   file list activelog /common/drf/localDevice/ detail
   

   Verify that drfComponent.xml is present in the output and has a recent timestamp corresponding to the service restart time.

5. Initiate a Manual Backup: Navigate to the Disaster Recovery System interface on the CUCM Publisher. Select the IM&P server in the backup device list and initiate a manual backup. Monitor the backup status for successful completion.

5.0 Post-Resolution Monitoring and Escalation

If the backup continues to fail after following this procedure, re-examine the drf_master.log and drf_local.log files for new or different error messages. The issue may be related to a specific component agent failing to register with the newly generated manifest. In such cases, collect a log bundle from the DRS interface (Cohesity DRF Logs) and escalate the case to the Cisco Technical Assistance Center (TAC) for further investigation.