This past week the NetApp A-Team met at NetApp’s facility in Research Triangle Park, North Carolina for our annual ETL meeting (Extract, Transform, Load). The 25 member team from NetApp partners, customers, and alliance vendors from around the globe, combined with another 15+ NetApp-employed A-Team tech advisors, spent two days in a room with NetApp product managers, technical marketing engineers, and members of the executive team to discuss NetApp’s future. We covered product roadmaps, gave feedback (the good, bad, and ugly) on our experiences from the field, and discussed various strategy initiatives that NetApp plan to deploy over the upcoming months. Unfortunately, I cannot share many of the technical details that we discussed during the week. Those items will come to fruition over the coming weeks and months. However, I will discuss an overarching theme from the week:
Storing data effectively, efficiently, and safely has become a commoditized business.
There are dozens of manufacturers that can do this, each with a slightly different approach. But overall, the value and objectives that they bring to their customers are the same. They each provide storage platforms with data efficiencies (i.e. dedupe, compression, etc.), good performance through the widely popular use of flash technologies, and they all employ some form of resiliency (i.e. HA controllers, RAID, erasure coding, etc.). Given this, how does a company like NetApp differentiate itself?
Moving, analyzing, and orchestrating the use of data is still a challenge.
During NetApp Insight 2015, NetApp announced a new initiative called the Data Fabric. At the time, and for many months following, the idea of the Data Fabric was very ethereal. We understood the concept – allow data to seamlessly and easily move wherever it needed to be in a hybrid cloud architecture to get the most value out of the data. The execution and the ecosystem surrounding the Data Fabric was the biggest unknown. We had an inkling that it would probably involve some use of SnapMirror (NetApp’s primary replication engine) to allow data to easily replicate across on-premises infrastructure and cloud service providers. Beyond that, no one outside of NetApp really had an idea what the platforms would look like at the end-points.
Today, the story has evolved. Since Insight 2015, NetApp has further integrated the Riverbed Steelstore acquisition into AltaVault, purchased SolidFire, further embraced Object Storage with maturing StorageGRID, and they’ve made great strides regarding their partnerships with Amazon Web Services and Microsoft Azure. Each of these items play directly into the Data Fabric strategy in the following ways:
A number of years ago, NetApp had a product called NearStore. NearStore was a pared down version of a FAS with SATA disk drives that an organization could use as a target for SnapMirror and SnapVault in order to copy snapshots off of their primary storage array. It was also possible to use NearStore as a VTL (virtual tape library) for organizations wanting to leverage disks for backups but with backup products that did copy data management through interfacing with tapes. Once the need for VTL went away, NearStore was retired. Some of the functionality was integrated into FAS, but for the better part of a decade, NetApp really didn’t have a separate, cost-effective platform to handle copy data management for native NetApp snapshots. NetApp attempted to make a play at acquiring DataDomain back in 2010 but lost out to EMC.
Enter late 2014… NetApp acquires Riverbed’s Steelstore product (also formerly known as Whitewater) and rebranded it as AltaVault. The big difference between AltaVault and other backup target disk arrays is its deep integration with cloud storage. The economics around cloud storage (i.e. AWS S3, AWS Glacier, Azure Blob, etc.) have changed the landscape of backup copy data management. Why store all of that data on-premises when you can consume the storage as you need it from a cloud service provider?? AltaVault now has integration with SnapMirror. Without any other product or software utility, an organization can setup and replicate NetApp snapshots from their primary FAS/AFF disk array to an AltaVault for short- or long-term storage. This has incredible implications when you think about pairing AltaVault with 3rd-party backup tools like Veeam and Commvault that can orchestrate NetApp snapshot backups. Copy data management becomes incredibly efficient, cost-effective, easy to manage.
Some people think NetApp purchased SolidFire because they didn’t have a good strategy around the use of flash. That couldn’t be farther from the truth. By the time NetApp purchased SolidFire, their AFF disk arrays (the all flash and tuned versions of their FAS systems) were well on their way to breaking all sorts of performance benchmarks.
NetApp purchased SolidFire because it filled a specific use case – a platform for consumable storage in much the same way that an organization would consume storage within a cloud service provider’s ecosystem. SolidFire’s combination of maximum and minimum QoS settings and it’s deep integration with platforms like OpenStack and Docker make it the perfect fit for organizations looking for storage resources that are consumable via API. This way of consuming storage resources plays into another big movement within the area of data center IT. HCI (hyper-converged infrastructure) platforms treat storage in a very similar fashion. And I’ll pause on that note…. more to come later. 🙂
But how does SolidFire fit into the Data Fabric story? Today, SolidFire can tier to object storage platforms like NetApp StorageGrid, AWS S3, etc. More to come on that front as well…
Object storage is a key component to a hybrid-cloud architecture. It provides a scalable platform for storing unstructured data and corresponding metadata across a geo-distributed topology. In regards to NetApp’s Data Fabric, StorageGRID is the underpinning for so many other capabilities. AltaVault and SolidFire can leverage StorageGrid as their object storage tier. Moreover, the upcoming and highly anticipated FabricPools feature in DataONTAP will be able to leverage StorageGRID as a means to offload snapshot capacity consumption from All-Flash FAS disk arrays. Without StorageGRID, the Data Fabric story wouldn’t be as compelling. The roadmap for StorageGRID is rich and probably the most exciting of any of NetApp’s products.
AWS and Azure Integration
One of the biggest challenges of leveraging AWS and Azure resources is getting your data there. NetApp provides a number of ways to do this. This includes Cloud ONTAP (an instance of Data ONTAP running as a virtual machine within either provider’s platform), CloudSync (a subscription-based utility to seamless move data into and out of either platform), and other “next to the cloud” options including NetApp Private Storage (NPS). Cloud ONTAP and NPS options provide the ability to use SnapMirror to replicate data to and from the cloud.
NetApp is also working with AWS on some very non-traditional solutions. This includes the new CloudControl offering. CloudControl is a software-as-a-service offering that provides data protection services for people leveraging other SaaS services such as Office365, Salesforce.com, Workday, etc. CloudControl doesn’t have any direct integration with NetApp’s traditional product offerings, but it provides a much-needed solution for an often overlooked problem. Too many clients rely on SaaS providers to protect their data, however, the responsibility of data protection still lies in the customer’s hands.
In conclusion, the Data Fabric story is becoming much more clear. Customers that have traditionally only used NetApp’s FAS products really need to be looking at the rest of the ecosystem. I’ve seen too many clients solely comparing NetApp to other storage manufacturers and trying to compare them at that level. It’s a huge oversight. Storing data is easy and, dare I say, boring. Leveraging the data to create value and provide insights within an organization is now the challenge…. and one I look forward to tackling!