Troubleshooting Network Attached Storage (NAS) Issues (4035486)

This article provides guidance on troubleshooting potential network and performance issues for customers using a Network Attached Storage (NAS) device as the location for a Rapid Recovery repository.

Note

Verify all configurations and settings against the latest product documentation, including the Rapid Recovery Sizing Guide, to ensure accuracy and alignment with Quest’s support policies.

Understanding Repository Flow

Standard backups in Rapid Recovery occur based on the protection schedule. Data flows through the network, from agent machines to the Core, and finally to the NAS repository. The repository stores deduplicated snapshots across all protected machines.

Challenges with NAS Devices

Using a NAS device as a repository introduces additional complexity compared to Direct-Attached Storage (DAS) or Storage Area Networks (SAN). Troubleshooting includes potential issues with network bottlenecks and increased I/O operations.

1. Performance Considerations:

   • NAS devices are heavily impacted by network latency and bandwidth limitations, leading to reduced throughput and data transfer delays.
   • Use high-speed NAS devices with redundant Ethernet connections (e.g., 10GbE) for optimal performance.

2. Dedicated Use:

   • NAS devices should be dedicated solely to Rapid Recovery repositories. Avoid using them for file sharing or other tasks.

3. Device Quality:

   • Quest recommends enterprise-grade NAS devices to minimize downtime and performance issues.
   • Devices with redundant network interfaces and scalable storage options are preferable.

4. Turn Off Scrubbing:

   • Disable scrubbing on the NAS device, as it can cause DVM checksum errors in the repository, potentially leading to data integrity issues.

5. Optimize Metadata Placement:

   • Store metadata on a local, high-speed disk attached to the Core server. This ensures faster access and reduces network load. Backup data, accessed less frequently, can remain on the NAS.

Intermittent Failures

NAS performance issues often manifest as intermittent failures (e.g., dropped connections or slow transfer rates during high I/O activity). To address these:

• Schedule I/O-intensive tasks to avoid overlap.
• Reduce transfer speeds to mitigate network congestion.

Environmental Factors

NAS devices are sensitive to network load, concurrent users, and backup frequency. Key considerations include:

• Avoiding network saturation.
• Optimizing retention policies and backup schedules.

Unsupported NAS Configurations

Quest does not support NAS devices with built-in deduplication if this feature is enabled. Data integrity requires that stored data remains unchanged.

Cloud-Connected NAS Devices

Primary repositories should not reside on NAS devices tiered to the cloud due to performance limitations. Cloud storage is suitable for replicated cores but not primary storage.

Recommended Configurations

Small Environments (1-10 agents):

• CPUs: 4 physical cores (e.g., Intel Xeon E5620 or better).
• Memory: 8 GB.
• Storage: 1.2x data capacity; performance ≥150 MB/s.
• NIC: Teamed 2x1Gb.

Medium Environments (10-20 agents):

• CPUs: 6 physical cores (e.g., Intel Xeon E5650 or better).
• Memory: 16 GB.
• Storage: 1.2x data capacity; performance ≥250 MB/s.
• NIC: Teamed 2x1Gb.

Troubleshooting Steps

1. Verify Network Access:

   • Confirm the NAS is accessed over a network path, not as direct-attached storage.

2. Inspect Event Logs:

   • Use Windows Event Viewer to identify connection or network issues.
   • Check the following logs:
     • Windows Logs > System
     • Common Network Issues for CIFS Repository:
       • Source: Mup, ID: 50, 139, 140
         • Description: {Delayed Write Failed} Windows was unable to save all the data for the file \\IP-Address\Offsite\Repo1\dfs.bmap. The data has been lost. This error may be caused by network connectivity issues.
       • Source: mrxsmb, ID: 50
         • Description: {Delayed Write Failed} Windows was unable to save all the data for the file \\SERVERNAME\MetaData2\dfs.bmap. This error may be caused by hardware or network failures.
       • Source: e1rexpress, ID: 27
         • Description: Intel(R) I350 Gigabit Network Connection #2 Network link is disconnected.
       • Source: b57nd60a, ID: 4
         • Description: Broadcom NetXtreme Gigabit Ethernet #2: The network link is down.

3. Restart and Check NAS:

   • Reboot the NAS to clear cache and restore connectivity.
   • Perform repository checks after rebooting.

4. Examine Hardware and Network:

   • Ensure NIC drivers are up to date.
   • Investigate potential network saturation or faulty NAS hardware.

5. Modify Transfer Settings:

   • Reduce Maximum Transfer Queue Depth by ~25% to alleviate I/O stress.
   • Navigate to Core Console > Machines tab > Configuration > Transfer Settings to adjust.

Frequently Asked Questions

Q: Why does my NAS drop connections?

• Overloaded I/O requests or defective hardware can cause dropped connections. Dedicate the NAS to repository use and verify its hardware.

Q: How can I diagnose NAS performance issues?

• Run single and multiple backup jobs to identify I/O bottlenecks.
• Create a secondary repository to isolate issues.

Q: What information should I provide to Quest Support?

• Issue details (e.g., dropped connections).
• NAS make and model.
• InfoGatheringTool logs and any configuration changes.