The answer is still no, and in this post I’ll talk about another key differentiator – scalability.
Scalability is an important issue because one of the common criticisms of hyperconvergence systems in general relates to how they scale. With our competitors’ systems, there’s validity to the criticism. But it doesn’t hold true for how Springpath delivers hyperconvergence, so this is a very important way in which the various hyperconvergence solutions are not created equal.
Let me back up. The whole idea of hyperconvergence is that you stitch a bunch of compute and storage nodes together and scale out your infrastructure on a modular basis, one node at a time.
The criticism comes from the fact that with appliance-based hyperconvergence solutions, there’s generally a 1:1 relationship between the compute capacity and the storage capacity in the appliance. So as your company’s needs grow and you need to scale the system, you add appliances, meaning you continue to have that same ratio of compute and storage capacity.
If what you really need at the time is more storage, you’ll be paying for compute capacity you don’t need. If you really need more computing power, you’ll be forced to buy storage capacity you don’t need at that time. These vendors do offer some flexibility in the sense that they offer small, medium and large configurations of their appliances. But the 1:1 ratio of compute to storage capacity is still an issue.
With the Springpath Data Platform software based approach and support for Adaptive Scaling, the criticism about the 1:1 compute-to-storage ratio is irrelevant. We don’t sell appliances, and our software supports whatever combination of compute, SSD caching, and storage capacity you need.
Unlike other hyperconvergence solutions, with Springpath you have complete flexibility in how you architect your system. Do you need a compute-intensive environment? You can run our software on a blade server environment and add SSDs to address performance demands. These can be independent decisions matching your business application needs. Do you need lots of storage? Create a rack environment with dozens of drives and run our software on that. You can mix and match blades and rack servers giving your applications the resources they need to meet business objectives. Springpath’s adaptive scaling gives you the flexibility to create the exact environment you need.
In addition to differentiating us from other hyperconvergence vendors, adaptive scaling is important because it makes our solution more affordable. One of our customers, Superior Document Solutions, runs Springpath Data Platform on a four-node cluster. Three of the nodes have both storage and compute. The fourth is a compute-only node that is used to run applications while leveraging storage off the other three nodes. By taking advantage of Springpath’s adaptive scaling in this way, Superior was able to put together an IT infrastructure that runs its entire business on a server based environment.
In the first two posts of this series, I discussed other important areas where Springpath differentiates from other hyperconvergence solutions: 1) Springpath is a software-only solution (vs. appliance-based systems), and 2) our underlying file system was built specifically for hyperconvergence (vs. the modified open source file systems of other hyperconvergence solutions).
It’s very clear that hyperconvergence solutions are not created equal. When you’re considering whether hyperconvergence is right for your company, I hope you’ll keep these issues in mind.