The future for storage admins is a topic near and dear to my heart. I was a storage admin, I helped build, teach, and write the curriculum and exams for EMC’s Proven Professional Program. What will happen to storage admins in the future?  With automation and cloud storage will there be an impact?

Digital Transformation is here again.

One of my first jobs at EMC was retraining mainframe SEs to attach UNIX servers to EMC storage. This was in another time of digital transformation when mainframe systems gave way to SANs and NAS systems that were attached to servers via SAN and IP networks (respectively). The applications of the day required this type of 3-tier architecture and a new IT job role – the storage admin- was required to manage the storage systems (see this blog post about that from 2002).

A new digital transformation is here, and once again applications are emerging that need a different sort of architecture than the 3-tier architectures that have been prevalent over the last two decades. Since the new architectures span clouds, will we need storage admins? I think we’ll need skills storage admins possess, but evolved to meet the new requirements of a multi-cloud world. Let’s talk about a few areas that seem perfect for storage admins to dig deep on now to prepare for the future.

CSI

CSI is the Kubernetes Container Storage Interface. It seems like Kubernetes is emerging as the favored de facto management platform for container-based applications. Containers are meant to be ephemeral, created and destroyed at will. Of course, anyone who has been a storage admin grimaces at this thought – what happens to the data?

Even more interesting, if you think about the new types of applications being created, what if I want to spin a container application up and access historical data to make decisions, to feed and train algorithms, etc.? Storage providers create CSI plugins for their storage systems, and the storage can be exposed via Kubernetes, without developers needing an understanding of managing storage. This will, however, mean that storage admins need to understand how the CSI driver shares information about the storage systems.

If I was still a storage admin, this is one of the areas I’d dive into immediately. CSI is very immature, and as this blog post on StorageGaga points out, persistent storage for Kubernetes gets hard as the applications grow. Not to mention the fact that CSI doesn’t have a distributed framework for persistent storage (yet). Understanding how CSI works to inform your organization will be valued (not to mention a blog post to explain it to everyone else as you figure it out!).

Cloud storage from traditional vendors

The traditional vendors know that they must provide a way for their customers to use their data on traditional arrays with cloud applications. At Cloud Field Day 8 last week, NetApp discussed their Cloud Volumes Service (their service that allows customers to consume NetApp storage natively in the public cloud and VMware). Here’s a quick demo of that service:

 

NetApp sees the future of the data center lining up like this:

  • 3 primary hyperscalers (Azure, GCP, AWS)
  • 1 primary hypervisor (vSphere)
  • NetApp Storage

This is a storage company betting that applications will be written across on-premises and public cloud instances. The skills of storage admins are going to be heavily leveraged here. If you’re wondering about the future of storage admins and thinking of next steps, it may be good to understand how storage companies like NetApp are building their next gen products.

Scale-Out File Systems and the Future for Storage Admins

In the public cloud, web apps are scaled out quickly as adoption to an application grows. This is done without sacrificing performance, many times using the cloud provider’s scale-out file system (under the covers of course).

At the moment, there are many on-premises scale-out systems that can extend to the public cloud (I previously wrote about Dell Technologies’ PowerScale). At Cloud Field Day 8, we heard from two scale-out vendors.

Ingeous concentrates on data management, which sounds like it wouldn’t matter to storage admins. But their platform built on a scale-out file system that can be used on almost any vendor’s NAS storage array.

Qumulo is a file data platform. They enable the ability to move data around to different platforms as dictated by the application needs. The examples they gave were content creation clouds, mortgage processing clouds, IoT clouds, and retail clouds. They are the workhorse behind HPE and Fijutsu’s scale-out solutions.

Understanding how scale-out systems work is also an important skill for storage admins. As applications are modernized, the data may shift to one of these types of file systems to get a handle on the data management.

Real Talk for the Future of Storage Admins

So what is the future of storage admins? That remains to be seen, but it is pretty clear that the skills storage admins have honed over the last twenty years are going to come in handy. If you’re a storage admin, deep dive into one of these domains or cloud storage and get ready to ride the digital transformation wave!