Last week I had a good experience attending IBM Edge 2014 where I had the opportunity to present a demo on IBM Enterprise Content Management (Filenet) over IBM Elastic Storage - the next big thing per IBM Storage strategy. The conference provided a good exposure to IBM Storage direction and strategy. One of the highlight of the conference was Jamie Thomas (GM, IBM Storage) keynote where she announced IBM Elastic Storage as one of the IBM Storage strategic initiatives. As stated in her keynote, IBM Elastic Storage is a step towards making the storage capabilities that powers IBM Watson (one of the mightiest IBM supercomputer) available for commercial usage, thus enabling the entire spectrum of business usecases to take advantage of it.
There were many good keynotes - one of them which resonated very well was the on-stage discussion between Arvind Krishna ( GM, IBM Dev) and Mike North (Sr. Director of Programming for the National Football League (US)), which interestingly discussed on the challenges of formulating entire season schedule for NFL such that the schedule is in favor of players, teams, sponsors, tv channels and the target audience. It's a classical example of big data problem and how IBM software and storage is applied to help get an optimum schedule for NFL - something only likes of IBM can deliver.
Highlight of other keynotes can be found here
Visit at the solution center exhibition was also very interesting. It looks like an industrial expo where IBM displays its best of product, solution with partners and innovation. At the solution center I hosted a demo over IBM Elastic Storage and ECM (Photo Attached below) along with my colleagues.
Since my focus area at Edge was around Elastic Storage, let me discuss on some key points around it:
- Elastic storage at the core of it is based on IBM GPFS but goes much beyond it.
- It's key design principle is "all in the software", thus truly qualifying it as software defined storage.
For example, ability to have native software defined RAID(GPFS GNR), ability to inherently host object stores like Openstack SWIFT (which in itself by design is software defined and agnostic of underlying hardware) as well as having NFS support on the same file system namespace as that of the object storage leads it to a converged storage.
- The vision of elastic storage does not stop here - it promises with inherent support of OpenStack (Cinder and Manila component), policy based WAN caching support through AFM (GPFS Active File Management), addressing analytic uses cases by leveraging file placement optimizer etc.
- It will also feature with capabilities that enable seamless file tiering to public storage clouds ensuring that the data at rest as well as data in motion is well secured and encrypted.
So in summary, this disruptive storage will help address what you have always longed for, with it's core design principle based on software defined environment. While this offering rolls and grows, today one can leverage the capabilities of IBM GSS storage to address your need of elastic storage.
Video on Elastic storage displayed at the POD
Following is a high level view of IBM elastic storage.
Me at the POD with my colleague from IBM Research