Configuring an NFS datastore for ESXi is simple using the NetApp OnCommand System Manager interface. I won’t get into the details of NFS vs iSCSI in this post but it is something I am evaluating. What I have learned about NFS is there is no multipathing. Instead, I have teamed my 10GbE interfaces on my controller to the switch to form one VIF. On the switch I am then using LACP. My other controller has it’s 10GbE in a VIF configuration too and both serve as partner interfaces for each other.
This post will go over the configuration of the Volume and Export on a NetApp FAS2240-2. It is a dual-controller in a single chassis. After configuring the Export and the proper permissions for the NFS Export, you then add it to your ESXi hosts as a datastore.
To start, create a volume for NFS. Select NAS as the Storage Type and configure your Volume size. In this example I am not using Thin Provisioning.
Moving on to Exports, you will notice an Export was automatically created from the volume previously provisioned.
In order for ESXi to mount this datastore we need to modify permissions for the export to allow the ESXi host to have root access:
Add root access to the ESXi host
That’s it. You now have an export ready to be mounted on ESXi.
NFS is a file-level in comparison to iSCSI, which is block-level. When you provision a virtual machine you will not have the option to select the virtual disk setting of thin or thick provisioning. This is because VMware doesn’t control this file system. It’s not running VMFS. The file system is handled by NetApp.
Are you using NFS or iSCSI and are you happy with performance? Let me know in the comments below.