NFS Shares

Path: Left sidebar > Storage > NFS Shares

The NFS page lets you create and manage NFS exports so standard NFS clients can mount cluster-backed file shares over the network.

NFS Shares

NFS Exports let you expose CephFS paths or RGW buckets over the NFS protocol to any standard NFS client.

What This Screenshot Shows: NFS page dashboard in a reference environment.

Prerequisite Check

Before creating exports, confirm:

• A file system exists in File System.

• An NFS cluster is deployed (Clusters tag visible at top of NFS page).

If no cluster tag is visible, NFS service is not running and exports cannot be created.

Important

If you see cluster tag but no exports, that is normal for new deployments. Click + Create Export to create your first share.

Actions in This Screen

Control	Action	Result
Clusters tag	Shows active NFS cluster	Indicates which cluster serves new exports
+ Create Export	Opens export creation panel	Configure and create NFS share

NFS Shares Table Columns

Column	Description
Export ID	Unique numeric identifier for the export
Path	The CephFS path or RGW bucket being exported
Pseudo	The NFS pseudo-root path clients use to mount this export
Cluster ID	The NFS cluster (ganesha daemon group) serving this export
Backend	Data source - CEPH (CephFS) or RGW (Object Storage)
Access	RW (read-write) or RO (read-only)
Protocols	NFS versions enabled (NFSv3, NFSv4)
Transports	Underlying transport protocols (TCP, UDP)
Squash	Root squash setting (no_root_squash, root_squash, all_squash)

Expanded Row

Export Details

Full export configuration in a two-column grid: FSAL type, CephFS filesystem name, user ID, security label, path, pseudo path, access type, and all protocol settings.

Client Blocks

Per-client access rules that override export defaults. Individual IP addresses or CIDR blocks can have different access, squash, and protocol settings.

Common Tasks

• Create An Export: Click + Create Export. Select the NFS cluster, choose backend (CephFS or RGW), set pseudo path and access type, then optionally add per-client rules.

• Edit An Export: Modify access type, squash, or per-client rules. Protocol and backend cannot be changed after creation.

• Delete An Export: Removes the export configuration. Underlying data is not affected.

Key Concepts

• Pseudo Path: The path a client mounts (example: /exports/data). Multiple exports can share the same NFS cluster but must use distinct pseudo paths.

• Squash: Controls how root on the client is mapped on the server. root_squash maps root to nobody (safer). no_root_squash preserves root access (trusted clients only).

• NFS Cluster: A Ceph-orchestrated group of NFS-Ganesha daemons. All daemons in a cluster share export configuration via RADOS-backed config storage for HA.

Action Decision Guide (Why, Impact, When To Act)

Action	Why User Does This	Impact	When To Do It
Create Export	Publishes filesystem path for client mounts.	Makes data reachable to NFS clients based on export policy.	Do this when clients need network share access to DFS data.
Edit Export	Adjusts access, squash, or per-client restrictions.	Policy changes apply immediately and can cause brief client reconnect.	Do this when security, compatibility, or access policy requirements change.
Delete Export	Removes an access path that is no longer needed.	Client access stops immediately while backend filesystem data remains.	Do this during decommission or when replacing export configuration.
Expand Export Details	Verifies effective settings and mounted-path mapping.	Reduces misconfiguration risk before troubleshooting or audits.	Do this after create/edit and during client mount failures.

How to Create an NFS Export

Use environment-specific values for cluster name, export path, pseudo path, client restrictions, and access controls. Available backend/source options can differ by deployment and enabled services.

1. Open Storage > NFS Shares.

2. Confirm cluster tag is visible.

3. Click + Create Export.

4. Select the Cluster used by your environment.

5. Select Storage Backend from the options available in your system. Examples include CephFS (file-system path export) or RGW (object bucket export).

6. Set the source path based on selected backend:

   • For file-system backends: select Volume and set export path

   • For object backends: select the target bucket/source exposed by your UI

7. Set NFS Protocol based on client requirements. NFSv4 is recommended for most environments.

8. Set Pseudo path using your chosen share name (must be unique in the cluster).

9. Set Access Type based on business need (RW or RO).

10. Set Squash mode based on your trust/security model.

11. Set transport protocol per policy. Keep both TCP/UDP only if required.

12. Optionally add Clients restrictions using approved IP/CIDR values.

13. Review your inputs, then click Create NFS Export.

Tooltips (Create Form)

• Backend: Use the backend options available in your deployment. Match the backend to your workload type and source data.

• Pseudo: This is the client mount path and must be unique in the NFS cluster.

• Squash: Choose the mode based on your trust model and access policy.

• Clients: Leaving this empty can allow any reachable host to attempt mount access.

Expected Outcome After Create

• Export appears in the NFS Shares table with an Export ID.

• Clients can mount using the configured pseudo path and protocol.

• Expanded row shows effective export configuration and client blocks.

What This Screenshot Shows: Create NFS Export panel in a reference environment.

Mount example:

mount -t nfs4 <nfs-server-ip>:<pseudo-path> /local/mountpoint

Create NFS Export - Field Reference

Field	Value / Options	Description
Cluster	Dropdown	Target NFS cluster serving this export
Storage Backend	CephFS / RGW	Data source backend for the export
Volume	Filesystem dropdown	Source filesystem to expose
Enable security label	Checkbox	SELinux-style labeling when required
CephFS Path	/ or subpath (CephFS), bucket source (RGW)	Backend path/source exported through NFS
NFS Protocol	NFSv4 (recommended), NFSv3 optional	Version compatibility setting
Pseudo	Example: /shared	Client-facing mount path; must be unique
Access Type	RW / RO	Write access policy
Squash	no_root_squash / root_squash / all_squash	Root/identity mapping behavior
Transport Protocol	TCP/UDP	Network transport options
Clients	Optional IP/CIDR list	Allow-list for client mount sources

Squash Settings Reference

Squash Value	What It Does	When to Use
no_root_squash	Client root retains root privileges	Trusted internal admin/backup hosts
root_squash	Client root mapped to unprivileged identity	Shared or partially trusted clients
all_squash	All users mapped to unprivileged identity	Highly restricted/untrusted environments

Warning

Leaving Clients empty can allow any reachable host to mount the export. For sensitive data, always configure explicit client restrictions.

How to Edit an NFS Export

1. Click row pencil icon, or expand row and click Edit.

2. Modify fields such as Access Type, Squash, or client restrictions.

3. Click Save.

Tooltips (Edit Export)

• Access: Use RO for read-only consumers and RW for writers.

• Client Blocks: Per-client rules can override export defaults for specific IP/CIDR entries.

• Squash: Choose based on trust model and least-privilege requirements.

Changes apply immediately; clients can briefly reconnect.

Expected Outcome After Edit

• Updated policy is applied to the export configuration.

• Connected clients can briefly reconnect and then continue access.

• Export row and expanded details reflect the new values.

Note

Protocol and backend cannot be changed after creation. Recreate the export if those must change.

How to Delete an NFS Export

Warning

Deleting an export causes immediate access loss for connected clients.

Before deleting:

• Notify users and request unmounts.

• Verify active client usage where possible.

Steps:

1. Click row trash icon.

2. Confirm deletion.

Tooltips (Delete Export)

• Delete removes only export configuration, not underlying CephFS or RGW data.

• Confirm client communication before deletion to avoid unexpected access interruption.

After deletion:

• Export is removed immediately.

• Connected clients lose access.

• Underlying filesystem and data remain intact.

Client recovery commands for stale mounts:

umount -f /mountpoint
umount -l /mountpoint

Export Details Panel Reference

Expand row chevron > to view full export configuration.

Field	Meaning
Export ID	Auto-assigned numeric export identifier
Cluster ID	Serving NFS cluster
Path	Internal filesystem path being exported
Pseudo	Client mount path
Access Type	Read-write or read-only policy
Squash	Root identity mapping mode
Protocols	Enabled NFS versions
Transports	Enabled transport protocols
Storage Backend	Backend data source (CEPH or RGW)
Filesystem	Source filesystem volume
Security Label	Whether labels are enabled
Clients	Allow-list (or unrestricted if empty)

Troubleshooting - NFS

Problem	Most Likely Cause	What to Do
No Clusters tag in header	NFS service not deployed	Deploy/verify NFS Ganesha via Infrastructure > Services
Cluster visible but no exports	Fresh deployment	Create first export
+ Create Export missing	Permission limitation or service absent	Check role and NFS service status
Client cannot mount	Firewall/path/restriction mismatch	Verify port 2049, pseudo path, and client allow-list
Mounted share is read-only	Access Type is RO	Edit export and set RW
Root permission denied	root_squash enabled	Change squash only for trusted clients
NFSv3 client fails	NFSv3 disabled or portmapper blocked	Enable NFSv3 and verify required ports
Stale mount after deletion	Client did not unmount beforehand	Run forced/lazy unmount commands
Pseudo path already in use	Duplicate pseudo path	Use unique pseudo path

Note

If issues persist, open a support ticket via Alerts workflow or Karios Support.