Storage Trends Part 2 — 3D Chess
I’m going to do something a bit risky and perhaps crazy. I’m going to perform a comparative analysis of various solutions in the storage market, and in the process risk starting a thousand vendor flame wars.
I hope and don’t think it will come to this, but still why would one want to do this? In this post I wanted to answer this as well as “set the table” for the actual analysis by discussing a few more issues around this topic (which will hopefully make a large Part 3 that much smaller).
DISCLOSURE: My current employer is both a NetApp Partner and a VMware Partner
The idea occurred to me last year after working with various storage technologies such as PernixData FVP and also researching other storage solutions ranging from All Flash, to hardware independent solutions like VMware VSAN and more. At first I wanted to write a blog post just on the PernixData (which I may still do) but when looking at the market and the new disruptive storage solutions it occured to me that a new paradigm was forming. In trying to find the optimal solution a few things became clear to me.
One is that the storage feature set is increasingly provided within the software and not the hardware – and the industry is converging somewhat to a common feature set that provides value (see Part 1 on Optimization Defined Storage).
Building on the above, we also have a new wave of hardware independent solutions with perhaps the most notable being VMware VSAN. Other trends include the increases use of flash storage, as well moving storage closer to the CPU (server-side).
The more I looked at various solution offerings the more the market looked to me like a three-dimensional chess board. Vendor X might have a clear advantage on one board, but on another board (or dimension) Vendor Y stood out. The more I looked at the playing ground, each chessboard represented a different value and/or performance proposition. For example, one solution category offers better CAPEX advantages while another offers better OPEX benefits.
This is key – the point of this exercise for me is NOT to compare and contrast vendors, but rather to identify what categories they fall into and what the nature of their value and performance benefits are for any given environment. The best solution will vary across budgets, workloads and existing storage investments – but by segmenting the market into different categories perhaps we can gain a better sense of where the different benefits exists so that we can understand the optimal value proposition and that we (hopefully) don’t find ourselves comparing apples to oranges.
Or in other words, storage is much like buying a house or a vehicle. What is the optimal solution for one family is going to be a different for another. Different budgets, different starting points and different needs.
There are a lot of exciting and disrupting changes taking place in the storage market. 2014 will be a very interesting year and we have more innovations on the way including leveraging host server memory, storage class memory and more.