Skip to main content
Pure Technical Services

vSphere Plugin User Guide: NFS Management

Currently viewing public documentation. Please login to access the full scope of documentation.


Prerequisites:  This KB article assumes that the remote vSphere Plugin has been installed to vCenter, that one or more FlashArray(s) have been added as a connection to it, that one or more ESXi host(s) or clusters have been configured for use with the plugin and that the FlashArray(s) are on Purity version 6.4.0 or later.

Recommended reading:

Creating an NFS Datastore

This feature was introduced in vSphere plugin version 5.3.1.

From the hosts and clusters view in vCenter, right-click (1) on the cluster or host where you want to mount the NFS datastore, hover over (2) Pure Storage then left-click (3) Create Datastore.


Select (1) NFS then left-click (2) NEXT.


Select (1) the cluster or hosts the datastore will be connected to then left-click (2) NEXT.


There are two options on this page: (1) Create a new file system or (4) Use a file system that has already been configured for NFS. If creating a new file system, populate the (2) Name and (3) Datastore Size then left-click (5) NEXT.


Optional: Select (1) Use a file system that has already been configured for NFS, select the (2) File System needed for this datastore and left-click (3) NEXT.NFSDatastoreCreation4-1.png

Select a (1) Pod or None for the file system to live in on the FlashArray. Pod membership is not a requirement of creating a file system on FlashArray.

A pod is a FlashArray object used to more easily manage block volume(s) or file system(s); currently block volumes and file systems are not allowed in the same pod. Pods are required for the use of ActiveDR, can have different file system policies applied to them, can have protection groups added to them and can be used to configure ActiveCluster. More detail on pods can be found in the FlashArray user guides.

Left-click (2) NEXT.


Select a (1) Policy type for the file system to have. Otherwise select Unrestricted or Restricted then left-click (2) NEXT.


Optional: If creating a custom policy is desired, populate a (1) Policy Name first. Select the (2) Client, Access and Permission required. Left-click (3)+ ADD RULE then finally left-click (4) NEXT.

For NFS datastores on VMware, no-root-squash is required and read-write is recommended for everything except if using the datastore for manual FlashArray snapshot recovery. More details on policies can be found here.


Check or optionally uncheck the (1) Enable Autodir checkbox then left-click (2) NEXT. It is recommended to enable automatic directory mapping in order to gain VM granular reporting and management on the FlashArray. More can be read about this feature in this KB.


Review the details then left-click (1) FINISH.


NFS Datastore Details

This feature was introduced in vSphere plugin version 5.3.1.

To see the FlashArray details of the NFS datastore, left-click the (1) Datastore view, left-click the (2) FlashArray-backed NFS datastore, select the (4) Monitor tab, select (5) Summary under the Pure Storage heading and finally (6) see the FlashArray details of the NFS datastore.


NFS Edit Datastore

This feature was introduced in vSphere plugin version 5.3.4.

This gives the administrator the ability to:

  1. Modify the datastore name
  2. Enable/disable autodir
  3. Modify the size of the datastore
  4. Modify the units of size of the datastore

Left-click on the (1) Datastores view in vSphere, right-click the (2) NFS Datastore you wish to edit, hover over (3) Pure Storage and left-click on (4) Edit Datastore.


In the window that pops up, you can modify the (1) Datastore Name, (2) Enable/disable Autodir, modify the (3) Size of the datastore and the (4) Unit for the size.

  1. In general, Pure Storage recommends enabling autodir on all NFS directories. Autodir allows the ESXi hosts to create, remove and modify the VM directories in the NFS mount directly. This enables VM granularity on the FlashArray for VMs inside of that NFS mount. More information can be found in this KB article.


NFS Destroy Datastore

This feature was introduced in vSphere plugin version 5.3.4.

In the (1) Datastore view of the vSphere UI, left-click on the (2) NFS Datastore to be destroyed, hover over (3) Pure Storage and left-click (4) Destroy Datastore.


In the window that pops up there is an important message to review. Because the NFS mount used by the NFS datastore will be destroyed on the FlashArray, it is important to consider if there are any other devices connected to that NFS mount that will lose access if the datastore is destroyed. After confirming the impact of the NFS mount destruction, left-click (1) DESTROY to destroy the NFS mount and the associated NFS datastore in vSphere.


NFS VM Undelete

This feature was introduced in vSphere plugin version 5.3.4.

In order to recover a VM that was deleted through the vSphere plugin, there needs to be a FlashArray based snapshot that the VM was part of in the past.

In the Datastore view in vSphere, right-click on the (1) NFS datastore the VM previously lived on, hover over (2) Pure Storage, and left-click (3) Undelete Virtual Machine.


Select the (1) Virtual Machine name to recover and left-click (2) NEXT.NFSVMUndelete2.png

Select the (1) Snapshot to recover from and left-click (2) NEXT.NFSVMUndelete3.png

Select the (1) Host the VM will be assigned to and left-click (2) NEXT.


Validate the configuration and left-click (1) FINISH.


NFS VM Point in Time Recovery

This feature was introduced in vSphere plugin version 5.3.4.

This will overwrite the VM for point in time recovery with a previous version of it.

From the Hosts and Clusters view in vSphere, right-click the (1) VM for point in time recovery (PiT), hover over (2) Pure Storage and left-click (3) Recover from Snapshot.


Select the (1) Source snapshot to overwrite the VM with and left-click (2) NEXT.


Note the (1) Warning text and details and left-click (2) FINISH to overwrite the current VM.