BlueXP Blog

12 reasons to use NKS

Written by Jonsi Stefansson | Nov 11, 2018 10:04:05 AM
Why is NetApp, a storage company, launching its own K8s service?

Well, firstly, I would like to emphasize that NetApp is and has always been a software company right from start when Dave Hitz, Michael Malcolm and James Lau founded the company with their game changing ONTAP Storage Operating System.

Secondly, we have launched Azure NetApp Files, the first Azure 1st party service, Cloud Volumes Service for GCP, 1st party service, Cloud Volumes Service for AWS, a 3rd party service, and we’ve got more coming. All those products run on K8s via our Service Delivery Engine (SDE) that basically runs on NetApp Kubernetes Services (NKS). The SDE is an application lifecycle management tool and a master helm chart that consists of all of our components, like our API server, Scheduler, Placement Policy Algorithm, Telemetric’s and Usage.

We have had preview customers running on Azure NetApp Files and Cloud Volumes Service for months to validate those services. We have been working very closely with those preview customers and found them to be extremely helpful and very influential in regard to what features we are exposing, service levels and just the overall development process. We owe them our utmost gratitude.

NKS lets you treat the world as your cloudWe really didn’t leave any stone unturned. We showed them how our services are running and how we are taking advantage of Kubernetes capabilities when it comes to autoscaling, self-healing and the zero downtime rolling upgrades, to name a few. We also showed them our SDE and NKS. It ended up with almost all of them wanting that service for their own homemade applications and to increase the rate of development in-house. The question “why aren’t you offering this as a service?” came up at every one of our workshops, so you could say that the decision was made due to popular demand.

I wanted to cover a little bit what makes NKS a very unique offering in no particular order: 

  • To begin with, we are dead-set on keeping true to the Cloud Native Concept and we have to due to our close relationship with Azure, GCP and AWS for our other services. Meaning that NKS can provision, maintain, and do a proper lifecycle management of your K8s clusters no matter where they are running.
  • NKS is a Pure Upstream K8s play and we support the latest release within a week. You can, however, choose to keep your cluster 1 release behind upstream for your application compatibility sake.
  • With NKS, you can Federate K8s clusters across multiple hyperscalers, and with us supporting NetApp HCI, we will be adding on-premise to the mix.
  • With the Stackpoint acquisition, we got a leadership role in Istio for Service Mesh, so you can distribute / route your application load across multiple clusters that you manage through NKS.
  • Application Lifecycle Management of the SDE is being ported to NKS in the next few months.
  • One of the hardest problem to solve for K8s, as reported on every single KubeCon, is and has been the storage, storage class and persistent volume claim.  Our integration of Azure NetApp Files and Cloud Volumes Service takes care of that automatically for you with Trident and provides you with unmatched flexibility of your underlying K8s storage. Since we are on all the hyperscalers (Azure, GCP and AWS), you can take advantage of the data management capabilities of ONTAP, like snapshots, clones, replication and migration.
  • NKS enables you to treat the world as your cloud with a single pane of glass management. No one else can provide this for you because they are not cloud native.
  • NKS is the only K8s service that allows you to run K8s on your own instances on all the public clouds where you control the master node, or on-top of GKE, AKS or EKS where you don’t control the K8s master node.
  • Simplicity, NKS makes it very simple to deploy and manage your K8s clusters.  It is hard work taking something this complex and making it this easy to consume. 
  • NKS fits everyone, from one developer to large enterprises with simple and transparent pricing model.
  • NKS is the only current K8s service that supports GPU containers.
  • NKS can be white-labeled and branded for our partners.


I am super proud of the team that started with having to build scalable services for the big 3 cloud providers, basically for ourselves that got this traction and demand from our existing customers, and we have added a large number of new customers that have signed up for NKS. In fact, since our launch a couple of weeks ago, 2282 clusters have already been deployed using NKS!  We at Greenqloud (acquired by NetApp 15 months ago) wrote the Service Delivery Engine and the Application Lifecycle Management on Kubernetes, but we always manually deployed and managed the Kubernetes clusters we were running on all the hyperscalers. With the Stackpoint acquisition and the launch of NKS, we have moved to using that purely for our SDE deployments everywhere. It used to take us 2-3 days standing up a new region on the public clouds. Now, with NKS, it takes us on average 23 minutes.

That is what you call eating your own dog-food …… or in this case drinking your own champagne ;)

I hope you take your time to visit cloud.netapp.com to start or continue your K8s journey with ease and comfort, backed with SLA’s and the good name of NetApp.

Want to get started? Try out Cloud Volumes ONTAP today with a 30-day free trial.