NetApp this week signaled a significant expansion of its cloud-native computing ambitions by acquiring StackPointCloud, a provider of a Kubernetes-as-a-service platform. Terms of the deal were not disclosed.
Anthony Lye, general manager for cloud data services at NetApp, says it’s already clear Kubernetes will become a de facto standard for running containerized applications on both public and private clouds. Via this acquisition, NetApp is now positioning itself to manage the entire Kubernetes cluster environment rather than focusing solely on storage, he says.
NetApp has been making contributions to the Trident project for creating persistent volumes on Kubernetes clusters for years. At the same time, NetApp has established technology alliances with Microsoft and Google through which both cloud service providers resell its software. The acquisition of StackPointCloud is a major step toward putting NetApp in a position to manage data as well as orchestrate storage and clusters across a hybrid cloud computing environment with the context of an integrated set of DevOps processes, Lye says.
NetApp is moving to make both infrastructure and data services more accessible to developers, Lye notes, adding there is no such thing as a storage administrator for a public cloud service. Instead, developers are building applications that programmatically invoke data via application programming interfaces. In time, that approach will be further extended to on-premises environments building private clouds, he says.
In fact, Lye notes that the rise of multi-cloud computing has put NetApp in a unique position to be able to orchestrate the management of data service across multiple platforms and cloud services. Traditional storage rivals continue to limit their focus to storage arrays deployed in an on-premises IT environment.
Lye notes most IT organizations still don’t fully appreciate the storage challenges inherent in deploying containers on Kubernetes clusters. Microservices made up of multiple sets of containers will be trying to access a common pool of data, resulting in significant I/O bottlenecks that will need to be addressed as more microservices-based applications are deployed in production. At the same time, containers assume the underlying storage resources can scale out on demand as the amount of data being accessed increases. The combination of those two factors will require IT organizations to embrace new approaches to storage based on Flash memory, he says.
NetApp is also anticipating that as Kubernetes clusters continue to mature, more organizations will wind up deploying them on bare-metal servers in place of virtual machines to reduce dependencies on virtual machines, which add a layer of infrastructure that in many cases will no longer be necessary. That doesn’t mean virtual machines are going away anytime soon. But Lye says NetApp expects there will be a need for storage platforms capable of supporting both bare-metal servers running Kubernetes and virtual machines.
It may take a while for that transition to occur. NetApp in the meantime is clearly betting it can play a much more strategic role in the age of cloud-native applications based on containers. In fact, Lye notes that NetApp has more than 600 developers writing software with the expectation that Kubernetes is about to radically transform multi-cloud computing.