NFS (Network File System) lets Linux systems share directories over a network. An NFS client mounts remote exports from an NFS server and accesses them like local filesystems. This guide covers setting up an NFS client on Rocky Linux 10, AlmaLinux 10, and RHEL 10 – from installing nfs-utils to persistent mounts and autofs.
We will install the NFS client packages, discover and mount NFS exports manually, configure persistent mounts via /etc/fstab, set up autofs for on-demand mounting, handle SELinux booleans, and troubleshoot common NFS client issues. The steps work on Rocky Linux 10, AlmaLinux 10, and RHEL 10 with nfs-utils 2.8.3 from the default repos. For the server-side setup, see our guide on installing and configuring NFS Server on Rocky Linux.
Prerequisites
- A server running Rocky Linux 10, AlmaLinux 10, or RHEL 10 with sudo/root access
- An NFS server with at least one exported share – see the nfs(5) man page for protocol details
- Network connectivity between client and NFS server (ports 111 and 2049 TCP/UDP)
- IP address of the NFS server (we use
192.168.1.10as the example NFS server throughout this guide)
Step 1: Install NFS Client Packages on Rocky Linux 10 / AlmaLinux 10
The nfs-utils package provides the NFS client tools including mount helpers, showmount, and the rpc.statd daemon. Install it with dnf:
sudo dnf install -y nfs-utilsVerify the installed version to confirm the package is in place:
rpm -qi nfs-utils | grep -i versionThe output should show nfs-utils version 2.8.3 on Rocky Linux 10 / AlmaLinux 10:
Version : 2.8.3Step 2: Discover NFS Exports on the Server
Before mounting, check what shares the NFS server is exporting. Use showmount to query the server:
showmount -e 192.168.1.10The output lists all available NFS exports and the clients allowed to access them:
Export list for 192.168.1.10:
/srv/nfs/data 192.168.1.0/24
/srv/nfs/backup 192.168.1.0/24If showmount hangs or returns “clnt_create: RPC: Port mapper failure”, make sure port 111 (rpcbind) and port 2049 (NFS) are open on the server firewall.
Step 3: Create Mount Point and Mount NFS Share Manually
Create a local directory to serve as the mount point:
sudo mkdir -p /mnt/nfs/dataMount the remote NFS share using the mount command with NFSv4:
sudo mount -t nfs4 192.168.1.10:/srv/nfs/data /mnt/nfs/dataVerify the mount is active with df -h:
df -h /mnt/nfs/dataThe output confirms the NFS share is mounted and shows the available space on the remote server:
Filesystem Size Used Avail Use% Mounted on
192.168.1.10:/srv/nfs/data 50G 12G 38G 24% /mnt/nfs/dataYou can also confirm the mount type and options with mount:
mount | grep nfsThe output shows the NFS version and mount options in use:
192.168.1.10:/srv/nfs/data on /mnt/nfs/data type nfs4 (rw,relatime,vers=4.2,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,timeo=600,retrans=2,sec=sys,local_lock=none,addr=192.168.1.10)This manual mount will not survive a reboot. The next sections cover persistent mounting.
Step 4: Configure Persistent NFS Mount with /etc/fstab
To mount the NFS share automatically at boot, add an entry to /etc/fstab. Open the file:
sudo vi /etc/fstabAdd the following line at the end of the file:
192.168.1.10:/srv/nfs/data /mnt/nfs/data nfs4 defaults,_netdev,rw,hard,intr 0 0Here is what each option means:
nfs4– use NFSv4 protocol (preferred over NFSv3 for security and performance)defaults– standard mount options (rw, suid, dev, exec, auto, nouser, async)_netdev– tells systemd to wait for network before mounting (critical for NFS)hard– NFS operations retry indefinitely if the server is unreachable (prevents data corruption)intr– allows NFS requests to be interrupted if the server goes down
Test the fstab entry without rebooting:
sudo mount -aVerify all NFS mounts are active:
df -h -t nfs4The output should list your NFS mount with filesystem type nfs4:
Filesystem Size Used Avail Use% Mounted on
192.168.1.10:/srv/nfs/data 50G 12G 38G 24% /mnt/nfs/dataStep 5: Set Up Autofs for On-Demand NFS Mounting
Autofs mounts NFS shares on-demand when accessed and unmounts them after a period of inactivity. This is useful when you have many NFS exports but only access them occasionally.
Install autofs:
sudo dnf install -y autofsEdit the autofs master map to define a mount point base directory:
sudo vi /etc/auto.master.d/nfs.autofsAdd this line to define /nfs as the base directory for auto-mounted NFS shares:
/nfs /etc/auto.nfs --timeout=300The --timeout=300 option unmounts the share after 5 minutes of inactivity. Now create the map file that defines which NFS exports to mount:
sudo vi /etc/auto.nfsAdd entries for each NFS export – the format is key options server:/export:
data -fstype=nfs4,rw,hard,intr 192.168.1.10:/srv/nfs/data
backup -fstype=nfs4,rw,hard,intr 192.168.1.10:/srv/nfs/backupEnable and start the autofs service:
sudo systemctl enable --now autofsCheck that the autofs service is running:
systemctl status autofsThe service status should show active (running):
● autofs.service - Automounts filesystems on demand
Loaded: loaded (/usr/lib/systemd/system/autofs.service; enabled; preset: disabled)
Active: active (running) since Sat 2026-03-22 10:15:00 UTC; 5s agoTest by accessing the auto-mount path – autofs will mount the share on first access:
ls /nfs/dataThe share mounts automatically and you can see the contents of the remote NFS export. After 5 minutes of inactivity, autofs unmounts it automatically.
Step 6: Configure SELinux Booleans for NFS
Rocky Linux 10, AlmaLinux 10, and RHEL 10 have SELinux enabled in enforcing mode by default. Several SELinux booleans control how NFS shares are accessed. Set the relevant booleans based on your use case.
To allow NFS-mounted home directories (for example, when /home is served over NFS):
sudo setsebool -P use_nfs_home_dirs onTo allow NFS exports with full read/write access:
sudo setsebool -P nfs_export_all_rw onTo allow NFS exports with read-only access:
sudo setsebool -P nfs_export_all_ro onVerify the boolean values are set correctly:
getsebool -a | grep nfsThe output lists all NFS-related SELinux booleans and their current state:
nfs_export_all_ro --> on
nfs_export_all_rw --> on
use_nfs_home_dirs --> onThe -P flag makes the changes persistent across reboots. Without it, the booleans reset to defaults on the next boot. For more on managing SELinux booleans and contexts, see our dedicated guide.
Step 7: Configure Firewall for NFS Client
The NFS client normally does not need firewall rules opened since it initiates outbound connections. However, if you have a restrictive outbound firewall policy, allow the NFS service:
sudo firewall-cmd --permanent --add-service=nfs
sudo firewall-cmd --permanent --add-service=rpc-bind
sudo firewall-cmd --permanent --add-service=mountd
sudo firewall-cmd --reloadVerify the services are allowed:
sudo firewall-cmd --list-servicesThe output should include nfs, rpc-bind, and mountd among the allowed services:
cockpit dhcpv6-client nfs rpc-bind mountd sshThe key ports used by NFS are:
| Service | Port | Protocol |
|---|---|---|
| NFS | 2049 | TCP/UDP |
| rpcbind | 111 | TCP/UDP |
| mountd | 20048 | TCP/UDP |
Step 8: Verify NFS Mounts
After configuring NFS mounts (manual, fstab, or autofs), verify everything works correctly.
Check mounted NFS filesystems:
df -h -t nfs4List all NFS mount details including options:
mount -t nfs4Test read access by listing files on the NFS share:
ls -la /mnt/nfs/data/Test write access by creating a test file:
touch /mnt/nfs/data/test-file.txt && echo "Write test passed" && rm -f /mnt/nfs/data/test-file.txtCheck NFS statistics to confirm the connection is healthy:
nfsstat -cThis shows client-side NFS RPC statistics. High retransmission counts indicate network issues between client and server.
Troubleshooting Common NFS Client Issues
Permission Denied When Mounting
If you get “mount.nfs4: access denied by server while mounting”, the NFS server is rejecting the client. Check these on the server side:
- Verify the client IP is in the server’s
/etc/exportsfile - Run
exportfs -raon the server to reload exports - Check that the exported directory exists and has correct permissions on the server
- Confirm the firewall on the server allows ports 111 and 2049
Stale File Handle Error
The “Stale file handle” error occurs when the NFS server export was changed or the server was restarted while the client had the share mounted. Fix it by unmounting and remounting:
sudo umount -f /mnt/nfs/data
sudo mount -aIf umount fails because the mount is busy, use lazy unmount:
sudo umount -l /mnt/nfs/data
sudo mount -aMount Hangs at Boot
If the system hangs during boot waiting for NFS mounts, the _netdev option may be missing from /etc/fstab. This option tells systemd to wait for network before mounting. Add it to your fstab entry as shown in Step 4.
Another option is to add nofail to the mount options so the system continues booting even if the NFS mount fails:
192.168.1.10:/srv/nfs/data /mnt/nfs/data nfs4 defaults,_netdev,rw,hard,intr,nofail 0 0NFS Version Mismatch
If the server only supports NFSv3 and you are mounting with nfs4, the mount will fail. Force NFSv3 in that case:
sudo mount -t nfs -o vers=3 192.168.1.10:/srv/nfs/data /mnt/nfs/dataCheck which NFS versions the server supports:
rpcinfo -p 192.168.1.10 | grep nfsThe output shows the NFS protocol versions available on the server:
100003 3 tcp 2049 nfs
100003 4 tcp 2049 nfsFor performance tuning and caching of NFS data locally, consider setting up FS-Cache for NFS on the client. For secure NFS access across networks, see our guide on configuring NFS with Kerberos authentication. NFS shares also work as persistent volume storage for Kubernetes clusters. For more on how the NFS protocol works at the kernel level, see the Linux kernel NFS documentation.
Conclusion
You have configured an NFS client on Rocky Linux 10 / AlmaLinux 10 / RHEL 10 with manual mounts, persistent fstab entries, autofs on-demand mounting, SELinux booleans, and firewall rules. The client can now access remote NFS exports reliably across reboots.
For production use, consider enabling Kerberos authentication for NFS (krb5p) instead of relying on IP-based access control, mounting with the noexec and nosuid options for security, and monitoring NFS performance with nfsstat and mountstats.
