A GlusterFS volume that accepts mounts from every reachable host can expose shared data beyond the intended client network. Restricting client access keeps new native mounts limited to known hosts or source ranges, which matters when storage and application networks are not fully isolated.
Native GlusterFS clients contact a cluster node to fetch the volume file before they connect to the bricks that serve data. The per-volume auth.allow option defines which addresses or hostnames may receive that mount configuration, while auth.reject can deny specific sources even when a broader allow rule would otherwise match.
Access rules apply to new mount attempts, so existing mounted clients may stay connected until they unmount or reconnect. Check the source address the cluster actually sees when clients connect through NAT, VPN, or proxy paths, and use TLS identities with auth.ssl-allow when certificate-based client control is the intended boundary.
Related: How to enable TLS for GlusterFS
Related: How to improve GlusterFS security
Tool: CIDR Allowlist Risk Checker
$ sudo gluster volume list volume1 volume2
$ sudo gluster volume get volume1 auth.allow Option Value ------ ----- auth.allow *
* means new native clients are allowed unless another rule or network control blocks them.
$ sudo gluster volume get volume1 auth.reject Option Value ------ ----- auth.reject none
auth.reject takes precedence over auth.allow, so a rejected client remains blocked even if its network appears in the allowlist.
$ sudo gluster volume set volume1 auth.allow 10.0.0.0/24,10.0.0.21 volume set: success
A wrong auth.allow value can block new client mounts until it is corrected from a trusted-pool node. Separate multiple entries with commas and no spaces, and prefer IP addresses or CIDR ranges when DNS consistency across storage nodes is uncertain.
$ sudo gluster volume get volume1 auth.allow Option Value ------ ----- auth.allow 10.0.0.0/24,10.0.0.21
Reset the allow rule with sudo gluster volume reset volume1 auth.allow when the volume should return to its default access behavior.
$ sudo mount -t glusterfs node1:/volume1 /mnt/volume1
$ df -h /mnt/volume1 Filesystem Size Used Avail Use% Mounted on node1:/volume1 1.0T 120G 880G 13% /mnt/volume1
$ sudo mount -t glusterfs node1:/volume1 /mnt/volume1 mount.glusterfs: failed to fetch volume file (Permission denied) mount: /mnt/volume1: mount(2) system call failed: Permission denied.
Denied native mount attempts are also visible in GlusterFS logs under /var/log/glusterfs.
Related: How to check GlusterFS logs