Oracle Cloud Infrastructure Documentation

NFS Client Reports an "NFS Server Not Responding" Message

An NFS client reports an "NFS Server not responding" message from a mount target.

This is a general troubleshooting guide that covers various factors related to this OS error message, which can have several causes. Review this information to help identify the cause and its associated troubleshooting tips.

The cause of this issue can be any of the following:

Cause 1: The NFS client itself can't communicate to the File Storage mount target IP address because of an OS- or Kernel-related issue.

Check the OS-related TCP network settings such as:

  • Whether the error message is coming from a single instance, or multiple instances.
  • Verify that you're using the latest kernel patch and nfs-utils package.
  • Check the load, performance, and memory usage of the instance during the affected period.
  • Verify MTU settings, in particular, check for any mismatch in the default MTU settings.
  • Check if iptables settings are causing dropped or denied NFS connections.

Cause 2: A TCP network communication (VCN) issue exists between the NFS client and the mount target.

Check for network communication issues:

  • Check whether any network firewall is causing dropped NFS requests.
  • Use rpcinfo -t <mount_target_IP> prognum to test for mount, NFS, and lockd port connectivity from the NFS client.
  • Use traceroute -n -T -p <NFS_ports> <mount_target_ip> to check connectivity.
  • Use nfsstat -o net to check for any dropped packets.
  • Use nfsiostat <mount_path> to check for detailed request data, such as retrans, RTT, and so on.
  • Use mountstats <mount_dir> to check for any abnormal RPC stats.

Cause 3: The File Storage mount target, which acts as the NFS server itself, isn't responding.

Check or collect the following for mount target issues:

  • Use dmesg -T to check for issues.
  • Check /var/log/messages for messages, including the timestamp of the message with timezone.
  • Collect the file system or mount target OCID.

In all cases, collecting a packet capture from the NFS client at the time of the error is helpful in troubleshooting the issue further. For example:

tcpdump -i <interface_name> host <NFS_client_IP> -w /tmp/FSS.pcap

For Windows users, collect information using Wireshark or similar tools.

Because these issues might be related to Compute instances and a VCN, issues can require collective troubleshooting. Engage OCI support and create a service request if required.