Heading for openness - looking at SDS virtualisation evolution from the reference of server virtualisation
In the previous three articles, we made it clear that the future of storage lies in the SDS technology. Breaking up the proprietary aspect of storage vendors by utilising a unified infrastructure to create a new open platform, which will eventually revitalise the storage industry that is currently walking towards its own destruction. Is this prospect a realistic one? Yes it is! The server market, over the last twenty years, has seen this kind of virtualisation take place and has successfully driven it from closed to open. What storage needs, is similar to the technology used in the VMware server arena, breaking up the current set-up and re-energising the value of servers.
Enlightenments from server virtualisation
Although today, VMware strides tall over the virtualisation arena, VMware was not actually the inventor of virtualisation. More than a decade prior to VMware’s founding, the server partition technology was already extensively used in mainframes and UNIX servers. Virtualisation on x86 was a technology that only gradually matured much later. In fact, if we analyse it purely from a technological angle, the level of virtualisation technology on modern AS/400 and AIX’s logical partitions is not inferior to that of VMware. But, everyone knows that the impact to datacentres from VWware was much greater than that of any virtualisation technology in the mainframes and Unix arena. Why is this? Because:
This is almost identical to the course of storage virtualisation! As of today, the virtualisation of the storage arena is still stuck in the dark period prior to the dawn of VMware’s founding. The ultimate goal of all of the mainstream virtualisation products on the market, such as IBM SVC and EMC VPlex, as well as HDS VSP’s virtualisation technology, is not to put somebody else’s product underneath their own platform or to tie the user up in their own closed ecosystem. This kind of technology is the same as the virtualisation of mainframe and UNIX, it cannot impact the entire storage industry.
Who will be the VMware of the storage industry?
Drawing lessons from VMware’s growth, I believe that the company should have several characteristics:
This is the only way it can be open and survive in such a hostile environment and eventually become a success.
Although hyper-convergence has, in recent years, garnered much attention in the storage arena, but I believe that starting from this point to completely upend the entire storage industry would still be extremely difficult. Because the goal of hyper-convergence infrastructure is to thoroughly replace extant storage array products. First, they would view all current storage arrays as the enemy. This needs a long period of evolution for a product to surpass the performance, functionality and stability of storage arrays, gradually eliminating the user’s countless worries, before it will be effective. This requires huge investments of time and money. Secondly, from multiple angles such as design, research and development, production and sales, the linked costs of hyper-convergence systems are not certain to be evident advantages over traditional storage arrays. Even if one assumes that all of these factors can happen, basic commercial norms of large investments, would expect sufficient returns. To best achieve the highest returns, would be to turn these hyper-convergence systems into another proprietary system. In this way, it would easily turn into another new, closed storage company, no different to today’s set-up, which is the exact opposite of opening up hyper-convergence.
So what is the answer to this question? It is the SDS virtualisation from 3rd party software companies!
Looking at the impact to the industry of storage virtualisation from the perspective of server virtualisation
I believe that FalconStor’s FreeStor solution can adequately meet the requirements above, and is a tool helping the storage industry walk out of the dark and back into the sunlight. First, as a 3rd party company specialising in software-defined storage, FalconStor has well developed co-operative relationships with current storage vendors; FreeStor extensively supports almost all storage products on the market, its high performance has already been verified by almost all suppliers including all-flash arrays vendors; the past 15 years of development history have already given ample confirmation as to its reliability and compatibility. It has rich management functions, of special note is the CDP disaster recovery protection technology which is already a worthy market leader. The newest direction of development for the technology is tighter integration with the OpenStack cloud platform, which would act as a bride helping corporations realise their move to a new framework.
The death of storage? Closing thoughts.
Using some combination like FalconStor’s FreeStor SDS virtualisation technology，the storage industry just might walk out from the shadows of night and back onto sunlit pastures. Of course this requires that the storage suppliers reassess their value. The high profit business model that was originally reliant on the vendor lock-in will soon be outdated. For the new wave, resistance is futile and it would be better to change oneself and welcome and enjoy it.
One day, storage virtualisation will finally become as mainstream as server virtualisation, look at the server vendors, their present is the future of the storage vendors. Mainframes and UNIX have already retreated to a small market share, and x86 servers, with the help of virtualisation technology rule the world. Although virtualisation greatly increases system utilisation, but there has been no reduction in the sales volume of x86 servers. On the contrary, they are steadily and healthily growing, because there are new applications and the emergence of new needs will increase the requirements of the market on servers. Most of the server vendors are still competing in the market. Is this what the storage market will be like in several years? Let’s wait and see.