Red Hat, Inc. announced earlier this year that their scale-out distributed file system solution, GlusterFS, could now support three primary modes of storage for OpenStack: block, file and object. The solution enables OpenStack cloud developers and operators to utilise GlusterFS to support storage modes. However, what is actually more interesting is the battle between experts; some say Gluster will transform the storage market, another says the software-based storage is the future, yet another analyst is far more uncertain and sees only limited use.
Ashish Nadkarni at IDC believes that Red Hat’s GlusterFS-based storage software will turn the market “on its head”, although he did admit that it may take a few years for this vision to be realised! In his recent podcast, he refers to Red Hat’s success in transforming Linux into an operating platform that can run as a cloud instance, as a virtualised platform, and could be a software-based operating system. When it comes to software-based storage, the premise is the same; a general purpose operating system that has been converted into a purpose-built storage platform. Nadkarni says that Red Hat will be able to commercialise GlusterFS, if their success with Linux is anything to go by, and the key to their success will be their market focus – the unstructured and semi-structured data that needs cheap, low-cost storage systems and platforms.
David Floyer, co-founder and chief technology officer at Wikibon is of the same mind as Nadkarni when he says that because software-based storage is the future for IT, with major industry leaders being involved in leading its acceptance in the marketplace, Red Hat’s GlusterFS is ideally placed to lead the way! Floyer believes that software-led storage will be based on commodity hardware and disks, commodity flash and controllers, and the software will be the differentiating factor. As Red Hat is part of the open source industry, it will be one of the key drivers, not to mention the possible cost savings, and then have a very good chance of making OpenStack hugely successful!
However, Marc Staimer, president at Beaverton, isn’t as confident and whilst admits that Red Hat has a chance of success, it will only be in the secondary storage market and for passive data, and is likely to find it difficult to achieve mainstream acceptance for their primary storage solution for real time data. Marc, in his podcast, agrees that GlusterFS will be a threat, but only to secondary storage, i.e. second tier of a storage system. As Red Hat is aiming to compete with the main storage players, a highly competitive marketplace, it will come down to the packaging and pricing. Whether Red Hat can repeat their Linux success, Staimer remains reserved; the open source Linux community was already a fairly established, stable group before Red Hat became involved; GlusterFS doesn’t have that following and it’s going to have to work hard to become more stable, with more functionality and more power, before that changes.
As they say, time will tell…