Intermittent NFS datastore mount failures during Cohesity restore operations

Products

• VMware vSphere ESXi

Issue/Introduction

Running a Cohesity restore operation fails during the initial mount stage with RPC timeout errors.

• Restore operations fail during the initial mount stage.
• Error message during operation may show: "Mount failed: Unable to get console path for mount."
• ESXi /var/log/vmkernel.log reports: "WARNING: NFS: 209: Got error 13 from mount call" or "SunRPC: Synchronous RPC cancel for client."
• The issue is intermittent and may succeed on retries or different hosts.

Environment

• VMware ESXi 8.x
• VMware vSphere 8.x

Cause

This issue is typically caused by network-layer packet drops or stateful firewall classification blocking SunRPC/Mount Protocol traffic (Program 100005) between the ESXi VMkernel interface and the Cohesity appliance. Secondary factors may include MTU mismatches (Jumbo Frames) causing packet fragmentation.

Resolution

To resolve this issue, perform the following steps:

1. Verify Firewall Classification: Ensure that the network security team has not implemented policies that drop "unknown-tcp" or RPC traffic on ports 111 and 2049.
2. Validate Port Reachability: From the ESXi shell, confirm connectivity to the Cohesity IP:
   nc -vz <Cohesity_IP> 111
nc -vz <Cohesity_IP> 2049
3. Audit MTU Path: If using Jumbo Frames (MTU 9000), verify end-to-end support with a non-fragmented ping:
   vmkping -I vmkX -d -s 8972 <Cohesity_IP>
4. Packet Capture: If the issue reoccurs, initiate a packet capture on the host storage interface to identify the point of failure:
   pktcap-uw --vmk vmkX --dir 2 -o /vmfs/volumes/<datastore>/nfs_issue.pcap

Additional Information

NFS mount failure or timed out on VMware ESXi: Error 13
Troubleshooting NFS datastore connectivity issues