There’s been a lot of speculation around what we’re doing at Nebulon…some spot on and some not so much. While we’re not ready to say just yet, I can absolutely tell you why we decided to start the company because in our case, the ‘why’ is just as important as the ‘what’. The best way to explain it is to take a walk down memory lane.
The CIO Challenge
When I was at 3PAR and later at HPE in the role of an HPE Fellow, I would often travel to see enterprise customers and walk through our architecture with various CIO’s. In one instance, I was at a F100 bank while our team was onsite setting up an all-flash array deployment. While the customer operations team was happy with the new features at the time, the CIO pulled me aside and asked me why he has to buy premium-priced external arrays when he has literally thousands of servers, each with a couple dozen slots for drives.
The answer, I explained, is that to provide the features he needed, I would have to put some piece of software on each server and he would hate me for it. That software would need to be different for different OS/hypervisors, would have to be installed and maintained on each server, and whose firmware updates and reboots would take storage offline—all while staff crossed their fingers that nothing would go wrong. I knew that was not the answer he wanted to hear.
However, that conversation stuck with me, and it took me some time (like a couple of years) to figure this out. After a number of conversations with other IT leaders, and frankly some timely industry developments, I was able to arrive at a more holistic solution to the CIO challenge shaped in large part by the following themes that came up over and over again:
Simple api-centric, cloud-based management
Savvy operations teams are eager for simple at-scale automation of many of the tedious and time consuming tasks associated with managing storage for core applications. A shift of the control plane to the cloud would make this possible, together with fleet management and automated software updates ‘out-of-the-box.’ And it would be the basis for moving to AI-based storage operations.
An alternative to the 3-tier architecture
Many organizations aspire to a move away from external storage and 3-tier architectures to the server-based, single-tier approaches in hyperscale data centers. These have proven to be cost-effective and easier to use for a specific set of workloads. All agree that array-based capacity remains the highest cost capacity in the industry, with the same SSD sold in a server at a fraction of the cost. But a move to a single tier approach requires a solution without the software footprint typical of the SDS/HCI approach.
A modern approach for critical workloads on-prem
Moving data to the cloud might seem like an obvious alternative for many organizations. But the reality is that for many CIOs, their most business-critical data assets, which often represent the crown jewel of the enterprises IP, often are held on-prem for a variety of reasons—including security, governance, cost, etc. So while organizations have plenty of storage options, there is currently no one solution in the market that combines an api-first, cloud managed alternative to legacy 3-tier architectures.
So what is the Nebulon solution? We will say more soon, but let me leave you with this one tidbit: what we have created is very disruptive and answers the CIO challenge by combining a few important shifts in the data center into one breakthrough offering.
And that, ladies and gentlemen, was how the idea of Nebulon, and more specifically, cloud-defined storage was born. Learn more about cloud-defined storage, and don’t forget to sign up to get a demo and join our beta program!
Side note, there was actually another reason we started the company. My co-founder & CTO Sean Etaati kept bugging me for years about doing a start-up, and I just wanted to get him out of my office. 🙂