I have read lots of interesting articles this week and wanted to compile them into a post, the theme of which is essentially where is the storage industry going? Storage used to be simple you had SAN, NAS and Object based storage if you were posh, on top of that you had your choice of vendor which was limited to a number of key players like IBM, HP, Netapp, EMC. Now the choice of storage vendors is mindboggling, with lots of smaller players coming to market with their own unique take on how the future of storage is going to look. The first article by Rich Castagna discusses just that, the bewildering array of architectures out there and the impact of flash on the market.
Next up is an article by Hu Yoshida at Hitachi which challenges the current trend for all flash arrays . In his article here he makes an interesting case that all flash arrays are a trend that will disappear. I tend to agree with him, in general only a small proportion of IOP’s will require flash performance and with nowhere for the less active or indeed dormant data to tier down, expensive storage is effectively being wasted. Also as currently flash drives have only been available in smaller sizes to their SAS counter parts reaching capacity requirements of a deployment is expensive. Whilst there will always be niche requirements for all flash, I think for the majority of use cases an array will benefit from a tiered storage system such as the standard 3Par models with AO or a hybrid array such as those from Nimble Storage . George Crump however disagrees, here he states his case for why flash arrays will become the default choice in the future .
In this article from The Register Chris Mellor discusses why the man behind the EMC ViPR software getting a more senior position EMC is important. The interesting part of the article is where he points out that with software defined products they are competing against their own traditional physical hardware such as the VNX.
Also in this week’s news DataGravity releases a hybrid array that is “data aware”. This means you can dig into your data from the array and see things like where data is being created most quickly and by who. Check it out in this YouTube video, it looks smart.
Going back to the first article it does make your headspin the number of different directions the storage industry seems to be going in right now. As with any technology I think over the coming years there will be a sort out and it will become clear which the great technologies are and those that promise lots and deliver little. At the heart of all this though I think will be good code and software will become king. This doesn’t necessarily mean that everyone will be going software defined but that the benefits the code of a storage product can offer will come to the front. Think about it HP didn’t pay $2.5 billion for 3Par because of its hardware capabilities they made the acquisition based on the quality and innovation behind the code.
The big winner from all of this will be the company that focuses on consolidation, taking the best in breed from all these innovative start-ups and combines them into reliable and winning formula.
Thanks to all the authors and Roger Luethy for as ever highlighting some of the most interesting storage articles from the week in his blog.
Follow 3ParDude on Twitter here