The Evolution of Hybrid Cloud Storage

Like many people, for the past few years I’ve been contemplating the vision of the hybrid cloud, and trying to understand the problems it could potentially solve.


The compelling reasons for deploying storage in an enterprise public cloud are obvious. First of all, it’s instant. I can easily create a cloud storage instance with a credit card and a few mouse clicks. Secondly, it’s disposable. After I create my storage instance in the public cloud, I can just as easily delete it and walk away with no cleanup whatsoever. Lastly, it’s inexpensive. I only pay for what I use, and there is no need to purchase equipment or build a data center. These are very compelling reasons indeed.


But, public cloud storage is not without its penalties. I must accept the fact that it’s OK to hand my data over to someone I have never met and trust that person will do the right thing with my data. I must also accept that the throughput and accessibility to my data will be limited by the infrastructure that my cloud provider has built. And, unfortunately, I must also throw all my existing data management processes out the window, because my cloud provider has an entirely different architecture with entirely different processes than I do.


Realizing the challenges our customers faced in cloud implementations, a few years ago, NetApp introduced NetApp Private Storage. This was an important breakthrough, as it allowed people to place NetApp physical storage systems in co-located data centers with big pipes directed connected to public cloud providers such as Amazon.


Blog cloud 1.jpg


NetApp Private Storage solved a couple of problems: It gave customers access to cloud compute services without giving up control of their data. It also allowed them to maintain higher levels of performance by purchasing storage systems that matched their desired performance levels. Finally, it gave customers the ability to keep their existing storage processes for things like provisioning, backup, and disaster recovery. But, NetApp private storage did not solve one problem – reducing the cost of storage in the public cloud.


To address the cost issue, NetApp recently announced Cloud ONTAP in conjunction with Amazon. Cloud ONTAP leverages the management capabilities that NetApp has developed over the past 20-odd years, but uses a virtual instance of Data ONTAP running within AWS.


Cloud ONTAP is a key component of NetApp’s hybrid cloud strategy.  It is a virtual machine version of Data ONTAP that is designed to run on top of Amazon EC2 compute instances. EBS storage volumes are assigned to the compute instance to create the equivalent of a Data ONTAP storage node that operates in the cloud.  Cloud ONTAP is an Amazon Machine Instance (AMI) that, via OnCommand Cloud Manager, can be deployed for quick delivery within the AWS marketplace, using Amazon's well known low cost pricing model.


Blog cloud 2.jpg


Like NetApp private storage, Cloud ONTAP gives our customers the ability to keep their existing storage processes for things like provisioning, backup, and disaster recovery. In addition, Cloud ONTAP utilizes Amazon EBS due to its higher performance and the ability to use guaranteed levels of performance from the underlying storage. Amazon S3 storage cannot provide these guarantees.

But unlike NetApp Private Storage, for customers willing to relegate their data to Amazon, Cloud ONTAP gives customers the pay-as-you-go cost benefit of public cloud storage.

Where is this all going? In my view, the most important thing that customers want is flexibility in how and where they store their data. Choice in public or private storage, choice in what public clouds they use, and the ability to change their minds and the choice they made in the placement of data.


Blog cloud 3.jpg


NetApp has spent the past several years developing a broad infrastructure of cloud partners. Our goal is not to build the world’s largest cloud or even to choose one cloud over another, but rather to coexist in all clouds and make it easy for customers to store, retrieve, and move data seamlessly between premises and clouds. I believe that over time, this will be the most successful path.


Resource: Creating Cloud ONTAP instances within AWS (7 minute video):


NetApp VP Phil Brotherton discusses NetApp Private Storage and Cloud ONTAP at AWS:reInvent 2014