Home > Storage > PowerFlex > White Papers > Using Dell PowerFlex NFS with VMware Cloud Foundation > NFS file system
There are a few best practices to follow when using NFS. Some are specific to VMware, while others are generic to NFS storage. In general, however, VCF requirements are going to map to the NFS best practices.
As NFS is IP-based network storage, it is essential that NFS traffic is separated from the other traffic like management in the VCF environment. Servers should use an MTU of 9000 or Jumbo Frames, to take full advantage of the bandwidth. In addition, a minimum of two 10 GigE NICs are required for VCF. If the environment supports larger networks like 25 GigE, which are present in this environment, it is preferable. VCF will “team” these NICs together when building the workload domain.
There are some VMware limits of note. By default, VMware restricts the number of NFS mounted datastores to eight for both NFS 3 and 4, however they do support 256 with a parameter change. Generally, Dell does not recommend changing the value unless it is necessary.
VMware controls both the heartbeating and locking of the NFS datastore. It has heartbeating parameters which VMware uses to check the health of the NFS 3 datastore. The mechanism is designed such that it allows recovery if a connection is lost for some short time period. The locking is also under the full control of VMware. The mechanism by which this is done does not need adjusting.