andyingram
on - last edited on
It’s been particularly interesting to observe what a mind-expanding concept the QFabric architecture is for most customers. They quickly grasp the benefits of a flat, any-to-any fabric to interconnect their data center infrastructure, and the concept of single-hop latency for all data flows while eliminating the challenges of managing the physical locality of processing and data is compelling as they virtualize their data centers. The ability to manage the fabric as a single, logical entity while eliminating the need to run multi-link protocols such as Spanning Tree or TRILL is almost too good to be true.
As I discuss the QFabric architecture with customers, the biggest challenge is getting them to fully understand the intricacies of the architecture. This is largely because they have been conditioned to view everything within the context of a network composed of individual switching elements. This is not surprising, given that every other vendor is taking this approach to building their data center solutions. To an audience inured to the complexity of networks, the QFabric solution is almost too simple.
The most important thing I tell them is that the QFabric architecture is not a network. It is a switch. Granted it is unlike any switch that has come before. It is faster, it is distributed, and it is more scalable than any Ethernet chassis switch ever built. Yet it retains the operational simplicity of a single switch.
What works best is to compare the QFabric architecture to the design and operational behavior of an Ethernet chassis switch, a well understood model. At some point the light comes on as the customer recognizes the elegance of the design. At this point in the conversation, I am typically asked one of my favorite questions: “Why haven’t other vendors taken this path?” It’s a good question, and there are two probable answers. First, it requires a major investment of engineering resources to design and build a proper fabric, and most vendors do not have—or don’t want to expend—those resources. The second reason they haven’t pursued this path is because a true fabric interrupts their existing business model. It would cause them to sell less, which is the natural outcome of simplifying any network. And selling less is not a compelling business model.
One other thing I’ve noticed about my visits with customers: the QFabric architecture is so innovative, so different, that describing it is like peeling back the layers of an onion. There are many dimensions and unique attributes that collectively make the QFabric solution such a differentiated architecture, and asking a customer to absorb all that information at once is a bit unfair. I’ve found it’s better to spread the discussion out over multiple conversations, giving customers time to reflect on and fully appreciate the new possibilities within their data center design and operation.
In each subsequent meeting, customers are ready to ask the next level of questions, and we peel back the next layer together. It is a very rewarding experience, and one that I am thoroughly enjoying.
