The Three Pillars of the Partnership - Engineering
Dive into the Engineering pillar that makes up one of the major pillars of the strategic partnership between Cisco and Nutanix
Building upon a previous post, the success of the Cisco-Nutanix partnership is measured by their ability to unify their respective strategies and deliver a differentiated solution to the market. These strategies are underpinned by three foundational pillars: engineering, support, and go-to-market efforts. This article will provide an overview of the engineering pillar.
I can't state this enough - coming to market with another "me too" product didn't make sense for either Cisco or Nutanix. The goal from the onset of the partnership was to take the strengths of both companies and combine them, but what does that actually mean? Let's dig in on the Cisco side first.
Cisco Engineering
Cisco Unified Computing System (UCS)
Cisco's x86 platform (UCS) offers a lot of differentiators in the market. One of these differentiators that other server vendors have been unable to replicate is to take what has typically been managed at the individual server level - such as firmware, BIOS settings, role base access controls (RBAC), etc. and manage those with software through an overarching piece of software called Cisco Intersight (or UCS Manager) to provide hardware level automation and consistency. This is done through pools, profiles, policies, and personalities. These concepts are covered fairly extensively from other sources but also apply to the Nutanix deployments.
The application of these profiles and policies ensures that all UCS servers for a given workload—in this case, Nutanix—are configured consistently for maximum performance and stability. When a Nutanix cluster is created, the Nutanix Foundation Central deployment tool interacts directly with a customer's Intersight instance to apply these precise settings automatically.
While Cisco UCS supports standard network interface cards (NICs) from vendors like Intel and Mellanox, the preferred and most effective approach is to use Cisco Virtual Interface Cards (VICs). This powerful piece of hardware virtualizes what would traditionally be physical network ports, including Ethernet and HBAs, through the same policy-driven framework. This was a critical co-engineering effort, as Nutanix had to adapt its platform to fully support this powerful feature. The ability to create virtual NICs (vNICs) on the Cisco VIC ties directly into the consistent, policy-based application across all nodes in a cluster.
Another major benefit to this approach allows Cisco to truly differentiate by allowing customers to use their hardware however they choose. One day, a Cisco UCS server can be running as part of a standalone ESXi cluster and the next re-purposed to be part of a Nutanix cluster. This hardware reusability function that separates Cisco from other server vendors deserves a dedicated article.
Fabric Interconnects
And how can we talk about Cisco UCS without bringing up fabric interconnects (FIs)? Integral to the Cisco UCS architecture are FIs. The integration of FIs with the solution provides numerous benefits, chief among them a highly stable, redundant, and high-throughput 10/25/40/50/100Gbps backbone network for storage traffic. Given that storage traffic in hyperconverged environments is often more intensive than in traditional converged setups, the network's importance is significantly elevated.
When including FIs as part of the overall solution, Cisco introduces a solution that truly includes not only compute and storage resources but the networking as well. Maybe we should call this "hyper-hyperconverged" or "maximum-hyperconverged!"
One common misconception with the Cisco solution is that fabric interconnects are required. This is not the case - FIs are not required! A customer deploying Cisco UCS C-Series rackmount Nutanix nodes have the choice of either using FIs or plugging into a pair of top of rack switches. This statement often leads to the next question: Does it have to be Cisco switches? No - it does not.
Cisco Intersight
Often overlooked with Nutanix deployments is the lack of seamless integration with the respective server vendor's management tools. When deploying a Nutanix cluster with other vendors, it creates isolated hardware clusters that typically only have visibility through the Nutanix Prism interface, lacking a true management tool integration. From the outset, Cisco was determined to avoid this challenge.
Cisco Intersight integration with Nutanix deployment is front and center to the overall solution and something that took a lot of co-engineering work. Why was this so important? When evaluating customer environments, it was critical to recognize that Nutanix clusters would not be the sole workloads within the data center. With this understanding, Cisco ensured that all Cisco UCS servers—regardless of whether they were running Nutanix or not—would be fully managed through its primary server management tool, Intersight.
This allows customers to have a single tool to manage their entire Cisco UCS estate and avoid creating islands of hardware clusters that are managed differently, providing consistency and true global visibility.
Nutanix Engineering
Nutanix Foundation Central
A subfunction of the Nutanix Prism Central (PC) software is called Foundation Central (FC). I had the "pleasure" to work with FC in my previous life. I was very excited for its potential but there were many aspects that were quite frustrating. One of which was the requirement for advanced DHCP scope options to be configured. I'll spare you the details and instead point out that Cisco has been the first server vendor to jump on board the FC train with both feet. The differentiation that Cisco brings with not only Intersight but the Device Connector software that resides on each local server allows deployments without advanced DHCP scope options. This was quite the leap for both engineering teams.
I can't sit here and claim that the process isn't without its faults (we'll walk through this in a future post) but because both companies agreed to integrate core tools, significant leaps were made to make the overall deployment experience with FC and UCS less frustrating.
If you take nothing else from this post: It's really ... really ... important to read the deployment guides.
Nutanix Lifecycle Manager
By integrating the Cisco tools, Nutanix also had to work with the Cisco engineering teams to ensure that none of the features and functions that have made the Nutanix software attractive for customers were degraded in any way. One of these features is the Nutanix Lifecycle Manager (LCM).
When utilizing Cisco UCS servers, the platform's Lifecycle Management (LCM) functionality is not only preserved but arguably enhanced. With Cisco, the Nutanix LCM engine now directly interfaces with Intersight to download the latest qualified firmware, ensuring customers are operating on the most current versions rather than a repackaged bundle provided through Nutanix.
The integration between Cisco's and Nutanix's core tools wasn't a happy accident but a conscious effort from the start of the partnership. All the blood, sweat, and maybe-not-tears have successfully culminated in a new, and more importantly, better and different solution for customers to modernize their datacenters.