An Introduction to SDS - Page 2
"If you are competing purely on the basis of software, and all the players have access to the same hardware, then you are forced to really innovate with what you can deliver in software. The X86 revolution has been a beautiful thing for innovation because vendors can no longer hide behind proprietary hardware. Innovation has to be done in software," explains Nadkarni.
So what kind of innovations can we expect? There's always the prospect of a new killer storage feature that no one has thought of before—but by definition it's not possible to give an example of what that could be.
More likely will be a push towards efficiency, as software vendors don't have a commercial interest in selling as much physical storage capacity as possible. "I think there will be a lot of focus on reducing the storage capacity we need," says ESG's Mark Peters. "Inline dedupe, compression, and other features that focus on better utilization of existing capacity. Software vendors will be quite happy to sell software that makes storage get used more efficiently."
Of course, companies won't be obliged to chooses a single storage software vendor for all their software needs—it should be possible to choose different software for different needs and apply it to different heterogeneous storage hardware as required.
This may sound like storage paradise but there are a few caveats that need to be considered. For example, do companies really want to buy storage hardware and then choose storage software separately rather than buy something integrated, and thus simpler?
IDC's Laura Dubois has her doubts. "If (companies) go the commercial or open source software route, they need to troubleshoot issues between the software stack and the hardware. The question becomes does the customer have the skill, and inclination, to perform this integration? For the SMB the answer is probably 'no' in the short run. However, for sophisticated enterprise accounts or at scale cloud providers the answer is likely to be different."
Just like software-defined networking, software-defined storage is already happening in a small way, through OpenStack, HP StoreVirtual, Nexenta, Symantec CFS and others. But it's likely that it will be several years yet before it hits the mainstream and existing storage hardware vendors have to really start to worry. That's because storage buyers are understandably conservative when it comes to new technologies and standards are yet to emerge.
There will also need to be a push to ensure interoperability, and software companies will have to figure out how to provide adequate customer service and support and customer service before businesses will commit their data to SDS products.
The big storage hardware vendors have been around for a long time and know a thing or two in this regard, so it's likely that they'll adapt rather than die. But whatever happens to them, SDS is here to stay, and it's going to have a material impact on the industry over time.
The good news is that it's storage users who will be the ones to see the business benefits.