Some of you might agree with me when I say NFS is a more than reasonable and acceptable solution for sharing files over the network.Traditionally NFS is considered to be slower than block based SAN systems because of higher latency introduced by the multi-hop networks and GbE network switches.
The client / server design of NFS makes it a stateless solution.Statelessness means that the server does not need to maintain state about any of its clients in order to function correctly.
True, different versions of NFS have their own intricacies and implementation complications and even security concerns, but over all, NFS version 3 has been a widely implemented and accepted file serving protocol since its invention by Sun Microsystems.
Storage companies like NetApp have adopted and implemented NFS in their products and even customized the protocol to address most of the complications found in earlier versions. This has paved the way for many other companies small and large, to adopt and implement NFS as their protocol of choice. Another reason why NFS has been successful is because every network operating system has had NFS ported to it in one form or another, and it is used in almost probably all Unix environment worldwide. It provides a convenient mechanism for sharing data across platforms, and is a robust, nearly ubiquitous solution to centralized data storage problems.
Engineers are familiar with it, users accustomed to it, and developers continue to improve it.
This last point is what I feel is the biggest win for the protocol. The brightest and most talented minds in the industry see value in constantly improving NFS and making it more versatile and scalable.
Nutanix, a next generation storage company, has implemented a very well thought out design to scale NFS to the next level for a virtualization environment. This distributed and parallel NFS implementation shares the high availability,scalability of the Nutanix SOCS architecture, which makes it hugely appealing.
In essence, Nutanix prevents the NFS master from becoming a single point of failure by distributing the Storage IO responsibilities among the NFS slaves. The NFS master manages all namespace functions, leaving data operations to the NFS slaves. When a VM does an IO operation, the NFS slave which is local to that VMs vmdk handles and completes the IO and returns a response.
NFS is a file-based protocol that avoids the overhead of VMFS. With NFS datastore now we can present the vmdk files directly to the VMs, similar to the concept of RDMs, but unlike VMFS where the IO calls would be written to the filesystem over block based LUNs.
NFS also has many other advantages over block level implementations of FC and iSCSI in Vmware.
Some of Nutanix NFS advantages are as follows :
The client / server design of NFS makes it a stateless solution.Statelessness means that the server does not need to maintain state about any of its clients in order to function correctly.
True, different versions of NFS have their own intricacies and implementation complications and even security concerns, but over all, NFS version 3 has been a widely implemented and accepted file serving protocol since its invention by Sun Microsystems.
Storage companies like NetApp have adopted and implemented NFS in their products and even customized the protocol to address most of the complications found in earlier versions. This has paved the way for many other companies small and large, to adopt and implement NFS as their protocol of choice. Another reason why NFS has been successful is because every network operating system has had NFS ported to it in one form or another, and it is used in almost probably all Unix environment worldwide. It provides a convenient mechanism for sharing data across platforms, and is a robust, nearly ubiquitous solution to centralized data storage problems.
Engineers are familiar with it, users accustomed to it, and developers continue to improve it.
Nutanix, a next generation storage company, has implemented a very well thought out design to scale NFS to the next level for a virtualization environment. This distributed and parallel NFS implementation shares the high availability,scalability of the Nutanix SOCS architecture, which makes it hugely appealing.
In essence, Nutanix prevents the NFS master from becoming a single point of failure by distributing the Storage IO responsibilities among the NFS slaves. The NFS master manages all namespace functions, leaving data operations to the NFS slaves. When a VM does an IO operation, the NFS slave which is local to that VMs vmdk handles and completes the IO and returns a response.
NFS is a file-based protocol that avoids the overhead of VMFS. With NFS datastore now we can present the vmdk files directly to the VMs, similar to the concept of RDMs, but unlike VMFS where the IO calls would be written to the filesystem over block based LUNs.
NFS also has many other advantages over block level implementations of FC and iSCSI in Vmware.
Some of Nutanix NFS advantages are as follows :
- Thin provisioning by default
- Provisioning is easy, no need to deal with fabric switches,storage masking,zoning etc.
- No filesystem overheard as in case of VMFS.
With Nutanix's version of NFS its now the perfectly logical and the wiser decision.
Now one can achieve the benefits of implementation ease, scalability, manageability, performance, and high availability, at a fraction of cost of NAS (network array storage) compared to companies like Netapp and eMC.
With NFS, the next generation Nutanix cluster blows away anything else in the market today, because not only has the design of this product considered and addressed the biggest pain points and vulnerabilities of NFS, it has also in the process created a robust scale on demand model of compute with 40-60% reduction in capital expenditure, while also providing a well proven and highly advanced shared-strorage functionality.
Did I hear anyone say...I want NFS !! Go to www.nutanix.com.
Now one can achieve the benefits of implementation ease, scalability, manageability, performance, and high availability, at a fraction of cost of NAS (network array storage) compared to companies like Netapp and eMC.
With NFS, the next generation Nutanix cluster blows away anything else in the market today, because not only has the design of this product considered and addressed the biggest pain points and vulnerabilities of NFS, it has also in the process created a robust scale on demand model of compute with 40-60% reduction in capital expenditure, while also providing a well proven and highly advanced shared-strorage functionality.
Did I hear anyone say...I want NFS !! Go to www.nutanix.com.
