A beginners guide to Nebulon Cloud-Defined Storage: Part I
This is part I of a multi-part series providing beginners a guide with the technical details surrounding Nebulon Cloud-Defined Storage.
Introduction
In July 2020 I announced that I was joining Nebulon as a Principal Tech Marketing Engineer. Changing jobs amidst COVID-19, let alone joining a pre-GA start-up amidst COVID is not for the faint of heart. However, the team and culture we’re building is nothing short of amazing so the transition was quite simple and painless. Since joining the team there has been a lot of learning and building going on behind the scenes getting the product and collateral ready for general availability (GA). I will say, one of my favorite parts of a start-up is having the opportunity to have such a profound impact on the product and its direction. I’ve even built my own YouTube show called #nerdtalk!
Nebulon Cloud-Defined Storage
Let’s lay the foundation of Nebulon Cloud-Defined Storage (CDS). CDS is a server-based, on-premises enterprise-class shared and local storage platform which consumes no host CPU, memory, or network resources. CDS is defined and managed through a secure cloud-based control plane, Nebulon ON, and an IoT device which runs in a customer’s application servers. The IoT device is OEM integrated at the time of server purchase (greenfield deployments) and delivered in the form of an integrated PCIe card, which is called a Nebulon Services Processing Unit (SPU).
Customers network-connect (10GbE or 25GbE) multiple SPU’s across on-premises application servers and compose them into logical groups, nPods. An nPod is one of the main building blocks for CDS and provides shared or local storage services for applications running on the servers. CDS is a platform which was built to be hypervisor, operating system, and container agnostic storage with data services (compression, deduplication, encryption, snapshot, etc.) traditionally these are features only available to external storage arrays. Unlike external storage arrays or hyperconverged solutions, CDS separates the data plane from the control plane. CDS is managed and monitored from a secure cloud-based control plane – Nebulon ON. This Software as a Service (SaaS) is accessible via web browser from anywhere on the planet and provides customers with deep insights for their global infrastructure with easily accessible AIOps, self-managed updates, and powerful programmability at any scale. Quite impressive!

Cloud-Based SaaS – Nebulon ON
CDS separates the data plane from the control plane. The data plan runs on-premises (customers maintain control of all their data) while the control plane runs in the cloud. This is a key architectural design principal which brings forth many benefits and poises Nebulon for success and scale into the future.
Simple, secure, and scalable tooling and instrumentation for operating a large-scale IT infrastructure should not be a complex burden for project teams or infrastructure administrators. There should be an inherent feature-set which does not require complex skills while providing automation, API and SDK capabilities with detailed infrastructure insights. Nebulon ON does just this.
Nebulon ON allows organizations to offer self-service infrastructure management for application owners. In addition, the cloud delivers Artificial Intelligence for IT Operations (AIOps) alongside centralized server, storage, and application management. This all enables the necessary tooling of real-time issue detection across the IT infrastructure stack and the facilities quick remediation for corrective measures. With global insights, Nebulon ON unlocks unprecedented opportunities for DevOps teams to control and query their servers, storage, and applications.
When the control plane is provided as-a-service, it is always up to date. Complex and lengthy update procedures are an artifact of the past. Users of CDS can make use of new features and enhancements instantly through rapid SaaS, cloud-delivered updates. Customers also acquire and maintain Nebulon ON like a SaaS in 1 or 3 year increments.
Where the rubber meets the road…Management and APIs
Nebulon treats the IoT endpoint, SPU, much like Google Nest treats the thermostat in your home – an API-first architecture which is controlled, managed and secured by the cloud. This design principle provides endless opportunities for automation.

To facilitate an easy to use, self-service provisioning process, the next building block, Nebulon Configuration Templates are applied to an nPod to deliver a consistent configuration to ALL servers within the pod. A Configuration Template is comprised of a few options: Storage – local, single host access or shared, multi-host access. Users also have the option to mirror data to backup nodes in the pod for additional data redundancy. Boot Volume Size, Boot Image URL and Additional Volume Sizes are options as well within a Configuration Template. Yes – I said boot image and URL! For example, customers may have a set of servers running a Linux distribution and then re-provision those same servers to run a completely different operating system. All of this is done by executing a script or playbook and rebooting the servers – all done remotely and via code.
Here’s an example of how to leverage Nebulon Configuration Templates from a December 2020 VMUG presentation I delivered.
Wrap up
Nebulon aims to be a stepping stone towards building the next generation self-service on-premises infrastructure. CDS’s API-driven architecture paired with having a physical IoT endpoint style storage controller within each application server or hypervisor host will allow for enhanced automation possibilities. This post covered a few of the building blocks of Cloud-Defined Storage, there is definitely a lot more to share!
Stay healthy, Clint
Pingback: A Beginners Guide to Nebulon Cloud-Defined Storage: Part II – Clint Wyckoff - Cloud Technologist