Redmonks Stephen O’Grady has an interesting post Software and EMC in which he discusses some of the problems that EMC, and other storage incumbents, are having with declining hardware sales and their change in focus to software and very specific high end hardware/software combinations (DSSD is one hell of an acquisition, everyone watching the storage space has been wondering what Bonwick et al have been up to, and Bechtolsheim has been obsessed with flash for quite some time).
Software and the Emergence of Commodity Storage
I would argue however that software has always been at the core of what EMC deliver though. What has made companies like EMC so profitable is their attention to areas such as availability and customer service. The age old story is that of the EMC engineer turning up to replace a failing component before you even knew it was a problem. Only software can facilitate that.
The storage market overall is growing, you just have to look at the uptick that companies like SuperMicro, Quanta and others are having, the entry and growing market share of multiple relatively new players such as Nutanix, Tintree, Tegile and others, and the incredibly growth of Amazon S3. Whats interesting here is that the growth is a combination of commodity components, and solutions for specific pain points.
This hardware growth is not where the real money is though and I think EMC see this, at least at an executive level, the real money is in managing the data and the exponential growth of the storage footprint at scale. Which leads us back to Stephens question
If it’s true that EMC is betting heavily on software to restore its hardware growth, the next logical question is whether this is the appropriate response
I don’t think EMC are trying to restore hardware growth at all, I think they are trying to completely shift their business model in the longer (5 to 8 years) term.
At the customer facing software, and developer, level we have removed many of the requirements for highly available hardware solutions – the high margin business EMC is the dominant player in – and moved to rapidly scaling solutions. This moves the real value in storage to managing and provisioning of storage, rather than supplying the underlying storage infrastructure (if you happen to provide it at a high margin, well and good, but its no longer your primary goal).
One point I do have to take up is around registration. I am all for ease of access, but industry context is important. Stephen notes
..the company keeps assets like ViPR locked up behind registration walls. In a market in which technology decisions are being made based more on what’s available than what’s good, that’s an issue
This is very much a cultural thing, traditional vendors such as EMC, IBM, Oracle and the like will always look for registration. EMC are not alone in looking for registration in the software defined storage market place either. If we look at true “software defined storage” vendors (not requiring a hardware product) such as Nexenta, OSNexus and StarWind only Nexenta provides a download that doesn’t require registration.
Bluntly put storage is hard, spinning rust breaks and the traditional way of controlling that is to have a defined hardware stack. When you are providing a software defined storage solution you run into issues similar to what Linux encountered in years gone by such as people installing on older hardware configurations with poor device support or broken components and so on. You then have to deal with the fall out from this on social media, forums etc. Registration allows you to understand a little about who the customer is, and understand if they are a potential sale worth investing time in, or someone (like me for instance) who has just downloaded the software to take a look. Its far from ideal, but its a commercial reality.
I would also argue that storage decisions are made on price, reputation and performance requirements rather than availability of a product to play with. Its a different market place, and DevOps are now driving some of the purchasing decisions. The capex required for storage is a massive proportion of most IT budgets, and any one dealing with operations and “spinning rust” will be happy to have a vendor involved in a proper proof of concept. Registration is not a barrier in that context.
Do I think EMC will succeed with ViPR? I am really not sure. As a preferred vendor to traditional customers, yes definitely. In new markets? Well their developer and devops story is still very hard to follow. Once you actually find their developer site you will see just how little activity there has been there so far, and the limited content is, to my mind anyway, still focused on very traditional storage admins.