How to export GlusterFS volume as NFS share

NFS clients often need one server endpoint even when the files behind that endpoint live on a distributed GlusterFS volume. NFS-Ganesha can serve the volume through the GLUSTER FSAL, so Linux, Unix, and appliance clients mount a normal NFSv4 path without installing the GlusterFS FUSE client.

NFS-Ganesha reads export definitions from /etc/ganesha/ganesha.conf and any included export files. The GLUSTER FSAL uses libgfapi to reach a trusted-pool node and named volume, while the Pseudo path defines the NFSv4 location mounted by clients.

A production export needs a stable gateway address or a separately managed high-availability design, restricted client networks, and root squashing unless a documented workload requires root ownership writes. Examples use Fedora or RHEL-family dnf package names when a repository provides nfs-ganesha-gluster; RHEL-compatible hosts may need an additional storage repository before that package appears.

1. Confirm that the GlusterFS volume exists and is started.

   $ sudo gluster volume info volume1
   
   Volume Name: volume1
   Type: Replicate
   Status: Started
   Number of Bricks: 1 x 2 = 2
   Transport-type: tcp
   Bricks:
   Brick1: node1:/srv/gluster/brick1/volume1
   Brick2: node2:/srv/gluster/brick1/volume1

   Run this from a trusted-pool node before configuring the NFS-Ganesha gateway.

2. Install NFS-Ganesha with the GLUSTER FSAL package on the gateway host.

   $ sudo dnf install --assumeyes nfs-ganesha nfs-ganesha-gluster
   ##### snipped #####
   Installing:
    nfs-ganesha
    nfs-ganesha-gluster
   ##### snipped #####
   Complete!

   If dnf reports no match for nfs-ganesha-gluster, enable the GlusterFS or NFS-Ganesha repository supported by that distribution before continuing.

3. Create a directory for dedicated NFS-Ganesha export files.

   $ sudo mkdir -p /etc/ganesha/exports

4. Open a new export file for the volume.

   $ sudo vi /etc/ganesha/exports/volume1.conf

5. Add an EXPORT block for the GlusterFS volume.

   EXPORT {
       Export_Id = 101;
       Path = "/";
       Pseudo = "/volume1";
       Protocols = 4;
       Transports = TCP;
       Access_Type = RW;
       Squash = Root_Squash;
       SecType = sys;
   
       FSAL {
           Name = GLUSTER;
           hostname = "node1";
           volume = "volume1";
       }
   
       CLIENT {
           Clients = 192.0.2.0/24;
           Access_Type = RW;
           Protocols = 4;
           Transports = TCP;
           Squash = Root_Squash;
       }
   }

   Keep Export_Id unique on this NFS-Ganesha instance. Pseudo is the path clients mount over NFSv4, and Squash = Root_Squash prevents client root from creating server-side root-owned files.

6. Open the main NFS-Ganesha configuration file.

   $ sudo vi /etc/ganesha/ganesha.conf

7. Include the volume export file near the end of /etc/ganesha/ganesha.conf.

   %include "/etc/ganesha/exports/volume1.conf"

8. Enable the nfs-ganesha service for future boots.

   $ sudo systemctl enable nfs-ganesha
   Created symlink /etc/systemd/system/multi-user.target.wants/nfs-ganesha.service -> /usr/lib/systemd/system/nfs-ganesha.service.

9. Restart NFS-Ganesha to load the export.

   $ sudo systemctl restart nfs-ganesha

   A restart interrupts clients already mounted through this gateway. Schedule the change or move clients before restarting a production export.

10. Confirm the nfs-ganesha service is active.

    $ sudo systemctl status nfs-ganesha --no-pager
    * nfs-ganesha.service - NFS-Ganesha file server
         Loaded: loaded (/usr/lib/systemd/system/nfs-ganesha.service; enabled; preset: disabled)
         Active: active (running) since Tue 2026-06-16 10:08:21 UTC; 8s ago
    ##### snipped #####

    If the service fails to start, inspect the NFS-Ganesha log and check the export file for an invalid Export_Id, unreachable hostname, or wrong volume name.

11. Allow the NFS service through firewalld on the gateway host.

    $ sudo firewall-cmd --zone=public --permanent --add-service=nfs
    success

    For an NFSv4-only export, client traffic uses TCP port 2049. NFSv3 exports require additional RPC services and firewall planning.

12. Reload firewalld to apply the rule.

    $ sudo firewall-cmd --reload
    success

13. Create a mount point on an NFS client.

    $ sudo mkdir -p /mnt/volume1

14. Mount the NFSv4 pseudo path from the gateway host.

    $ sudo mount -t nfs4 files1:/volume1 /mnt/volume1

    Install the client-side NFS utilities package first if the host does not have the mount.nfs4 helper.

15. Verify that the client mounted the export over NFSv4.

    $ findmnt -T /mnt/volume1
    TARGET       SOURCE         FSTYPE OPTIONS
    /mnt/volume1 files1:/volume1 nfs4   rw,relatime,vers=4.2,proto=tcp,sec=sys