Content Copyright © 2015 Bloor. All Rights Reserved.
“Flash-based storage systems are not just a storage issue. The increases in capacity, the reductions in cost, and the data management tools now available for them, are providing customers with a major business process re-engineering opportunity.”
That, at least, is the view of Chris Johnson, EMEA Vice President and General Manager of HP’s Storage Division, who sees the rapid growth of Flash storage systems leading directly to enterprise users starting to launch some major rip-and-replace activities with their existing data storage systems. He certainly sees significant growth in the use of Flash storage systems – the same technology as is used in USB stick memories and SD cards pumped up to systems offering Petabyte capacities and beyond – with around 500% growth in the last year.
While this is obviously starting from a low base, the signs of that rip-and-replace model getting underway is evidenced, he suggested, by a current and probably temporary dip in storage systems sales of around 12%. This is because Flash-based systems can deliver significantly greater storage capacity in a much smaller byte-for-byte form factor, and at a much lower byte-for-byte cost.
Johnson sees this trend accelerating as the traditionally embedded management software is extracted from the systems to become stand-alone, and far more comprehensive tools. The storage hardware then becomes a classic semiconductor-based commodity product, with all the implications of Moore’s Law waiting to be applied.
Part of the HP plan, therefore, is that the full management software suite, which should in theory be the company’s future ‘secret sauce’ in users developing new, re-engineered and re-architected business processes, will be available to work with any commodity storage, from Tier 1 Flash systems through to Tier ‘n’ near-line tape backup systems. This should also add greater flexibility when it comes to building not just Software-Defined Storage systems but the ‘Software-Defined Everything’ that will be the foundation for widespread business process re-engineering.
The combination of Flash storage systems, hyperconverged compute resources and Software-Defined Everything environments will mean that system architectures will become driven purely by the logic underpinning new business processes, not the need for resources to be located in specific physical locations.
This also moves the development and implementation processes out into the realm of the service providers. They, if they are any good at their job, should have a far deeper knowledge of the business practices and market needs of their customers – that is why they are important to global vendors like HP. Having Flash storage available as a service offering will be an essential start point, but the key will be in having access to that stand-alone management services suite. With that they are ideally placed to help customers not just migrate to the cloud but get some real business advantages as well.
Indeed, users have to start thinking about themselves as ‘software’ companies – in essence, businesses defined by the software they are using rather than the product or service they provide. In doing that they can then exploit more fully the capabilities service providers can offer. Most already offer customers infrastructure management services, but the trend is towards them offering full data management services, freeing users up to concentrate even more on their own core businesses.
The service providers signed up with HP will also, of course, have access to the latter’s expertise in using the management tools. According to Johnson, this is where HP’s ultimate justification for the troubled acquisition of Autonomy will really come into its own. Its capabilities in data management, such as controlling the exponential duplication of unstructured datafiles as they get shared will, he suggested, prove to be a major advantage for the company.
This is not all ‘wine and roses’ however, for there is an issue to be faced with Flash technologies further out. Flash won’t be the end of the development line in storage technology, which raises the issues of future incompatible formats and protocols. There is a growing need for a self-descriptive metadata standard so that all data can be transitioned to future data formats/storage formats, and remain readable.
There is not too much evidence of one emerging in the near future, however.