VMware has arguably done more to change how Data Center infrastructure is architected and built than any other company over the past 10-15 years. Today, most IT infrastructure leaders use VMware products somewhere within their organization.
A large proportion of that VMware infrastructure still runs on traditional 3-tier architectures or Converged infrastructure designs such as FlexPods and vBlocks. Over the last 10 years, Hyper-converged Infrastructure (HCI) systems like VMware’s vSAN and Nutanix have become a popular choice. According to analyst firm IDC, approximately $4-5bn is spent globally each quarter on Converged, Hyper-converged, and Integrated infrastructures.
Everyone looking to modernize their approach to VMware-based infrastructure typically has similar challenges and questions. How to eliminate infrastructure silos? How to ensure they are ready to support integration with the public clouds? How to run apps consistently? How to operate most efficiently? How to run the best and most flexible infrastructure stack without becoming reliant on any one vendor? How to ensure security and business resiliency? How to transform a traditional VMware environment into an internal cloud?
A smart new approach for VMware infrastructure
Nebulon was founded 3 years ago and offers a new and innovative approach for those considering the future of their VMware and Hybrid Cloud infrastructure. We provide a server-based approach to your VMware deployments without the overheads and OS restrictions of HCI, and 100% cloud-managed like public cloud infrastructure. This architecture provides you with the benefits of the public cloud while keeping your application data in your own data center. You can deliver smart services for VMware applications today, with the flexibility to also support containerized, or bare-metal applications as your infrastructure evolves.
Our customers have found that they can cut infrastructure costs in half compared to previous approaches. And because Nebulon doesn’t consume any server CPU, memory, or networking resources (unlike traditional HCI), they are able to use 100% of server resources for the VMware-based applications they run!
Already interested? Get started with Nebulon smartInfrastructure here.
1. Cloud-like VMware user experience
Many vendors talk about ‘cloud-like experiences’ for VMware infrastructure. However, infrastructure owners, managers, and users understand the reality is often somewhat different – whatever the marketing material and salespeople may say!
With traditional HCI infrastructure, management is usually done site-by-site or cluster-by-cluster. This makes software upgrades and provisioning extremely painful and time-consuming. All management console instances also need hosting, installation, configuring, and upgrading. Automation is cluster-by-cluster with APIs that can be at different revision levels. Not very ‘cloud-like’!
Nebulon offers a more modern, cloud-like approach for VMware – better for both IT management and VMware infrastructure end-users. Why believe our claims over those other vendors? Our smartInfrastructure architecture is fundamentally different. It is configured, deployed, and administered through our unique Cloud control plane, Nebulon ON, with centralized ML-based automation and a single cloud API across all clusters and sites. This approach dramatically reduces operational overhead, provides a simple API-first approach to infrastructure operations teams, and offers self-service VM provisioning for developer and application teams.
We find that 46% less time is spent on admin and reporting, with single click updates and reporting available – filtered by location, application, or internal group or project team. Soon, Marketplace templates will allow user provisioning of IT-certified VMware datastores in minutes, whilst also improving the consistency of your global infrastructure.
This blog and Powershell module demo by our Principal Solutions Architect, Aaron Patten explains these concepts in more detail.
2. Improved Availability and Security
Provisioning VMware infrastructure is more time-consuming and complex than it needs to be. This in turn can cause availability and security risks and is an issue because traditional HCI architectures manage data volumes, but are not able to control boot volumes and related configurations consistently. Also, HCI software updates and server OS/hypervisor maintenance takes the HCI node and critical storage resources offline, potentially placing your data at risk. Endpoint management is a tedious host-by-host, device-by-device approach, so critical SSD firmware updates are often neglected, creating potential security exposures.
Our unique Data Plane/Services Processing Unit (SPU) architecture provides rich data services and advances security models to improve data availability and minimize security threats. How? Nebulon software updates happen from a cloud control plane, Nebulon ON, delivering new features non-disruptively as-a-service. Software updates never take data/storage resources offline. The SPU & host server are in different fault domains, meaning server & hypervisor maintenance never takes data/storage resources offline. To further minimize security threats and availability risks, Nebulon provides for one-click, global SSD firmware, and bios updates.
Importantly, high-availability features and services are configurable, per Nebulon nPod cluster. This means that the right level of availability can be architected for a given VMware-based service or application. This also applies to container or bare-metal applications, should you require those as part of your modern data center architecture. Eliminate those infrastructure silos for good!
3. Lower Costs – Now and in the Future
There is no doubt that investing in HCI has improved efficiency and infrastructure utilization levels for many IT organizations. However, traditional HCI management overheads will always consume up to 25% of server resources, so one-in-four server and related software license purchases are still wasted. And, should you choose the re-deploy your infrastructure to container-based applications, pivoting to Kubernetes-based workloads still requires you to purchase a hypervisor!
Nebulon’s unique architecture and innovative pricing models will improve both your CAPEX and OPEX cost structures. This means you can demonstrate meaningful infrastructure cost reductions without compromising service levels. Benefit from at least 33% fewer server purchases and associated software licensing costs, as no server resources are consumed by data services in our architecture. The typical 1.25-1.33x density improvement means additional cost savings in data center space, power, and cooling.
If you are using VxRail or Nutanix for VMware today and are thinking about your move to containers and Kubernetes infrastructure, then Nebulon gives you a powerful alternative. With VMware, you need Tanzu. Nutanix has a similar approach with a containers stack on AHV. Building your VMware applications on Nebulon smartInfrastructure means VMware efficiency gains now, while you develop your K8s strategies. Nebulon supports bare metal & containers natively, avoiding hypervisor license costs altogether. This provides for further savings – typically 46% for native K8s deployments. You can manage VMware, Kubernetes, and bare metal infrastructure with same the cloud control plane—altogether more efficient, now and in the future.
VMware & Nebulon – Integrations
Nebulon provides support for VMware vSphere 6.7 and 7, with zero required drivers or extra configurations to worry about. We have developed useful integrations with VAAI, including features such as XCOPY. This off-loads cloning activity from the host back to the storage layer, so tasks like deploying new VM’s from templates are completed quickly and efficiently. Nebulon has also provided integration for vSphere DRS. No surprises there, but there are also unique capabilities that many of our customers find very useful indeed.
Ever tried and failed to find an accurate answer to this simple, but fundamental question – “What’s my VMware utilization across my estate?”. Our unique, cloud-controlled architectural approach means that we can provide infrastructure-wide views of VMware infrastructure and applications. This allows infrastructure and application owners to see live usage data and metrics. Grouping and reporting of these metrics are user-defined. Very different to an ‘already out-of-date standard PowerPoint presentation’, reviewed at the end of each quarter. Think of a Smart VMware Meter vs a Quarterly estimated VMware utility bill. Our COO, Craig Nunes, explains more about the fundamentals of smartInfrastructure in this recent blog post.
With Nebulon, all meta and performance data is collected and available to analyse. Nebulon requires no special drivers or special configuration to provide bootable & datastore storage for VMware. We provide guidance to VMware DRS to indicate of the most optimal layout of the VMs across the environment. We also integrate with vCenter telemetry and provide vSphere statistics; this means you can more easily and quickly troubleshoot and resolve any infrastructure issues.
Within an application template, we can specify that your preferred configuration of vSphere is being deployed. We provide this consistency of provisioning within Nebulon ON. Consistency drives reliability and tightens security. A template forces a level of standardization that enforces this consistency. With vCenter credentials, Nebulon ON records information about the vCenter environment, giving the much-coveted ‘single pane of glass’ management approach for global infrastructure deployment.
VMware & Nebulon – Applications
We find there are two types or groups of initial use cases for VMware + Nebulon smartInfrastructure.
The first is application-specific, such as end-user Computing (EUC, also often known as VDI/DaaS). Many organizations are already using GPUs to offload processing from the core Server CPU for EUC. An SPU offers a similar approach and advantages for the data plane, allowing you to protect precious CPU resources and maintain separate fault domains.
The second use case is for mixed VMware workloads – “the I/O blender”, as my esteemed colleague Martin Cooper calls it. This is infrastructure for your general purpose, and often unpredictable VMware workloads. This is where the majority of the $5bn infrastructure spend I mentioned earlier goes.
smartInfrastructure for VMware
VMware themselves have announced Project Monterey, an offload technology for storage and networking for disaggregated infrastructures that are now in an Early Access program stage for lab environments. However, this DPU/SmartNIC approach is only a part of the solution required to truly Modernize your VMware Data Center. See Chris Evan’s view here.
In summary, with Nebulon smartInfrastructure, you can access curated application templates that are continuously tuned to workload needs, efficient data reduction, and enterprise storage data services that allow VM admins to easily instantiate VMware clusters and operate them with ease—all at a fraction of the cost of external arrays and without the compute resource and related software license penalty of HCI solutions.
View our VMware configuration guide here.
For a more technical comparison of Nebulon smartInfrastructure compared to Nutanix HCI, click here.
For a more technical comparison of Nebulon smartInfrastructure compared to VMware vSAN, click here.