The storage industry has been undergoing constant innovation over the last several years with the advent of flash based arrays, automatic tiering, compression and de-duplication technologies. We believe in 2015, it will be “flash first” strategies, software defined storage and cloud technologies that will have the greatest impact on the storage market.
Prediction 1 – F1000 Enterprises Will Evolve to a “Flash First” Strategy
In 2014, a bevy of new all-flash array (AFA) products from EMC, NetApp, HP, IBM and Dell became mainstream products and joined privately-held companies Pure, Nimbus, Skyera and SolidFire in this rapidly evolving market. Even traditional hybrid storage vendors Nimble and Tegile introduced AFA versions of their products. Every storage vendor on the planet now sells AFAs. This is because they have proven to be truly cost-effective for Tier 1 apps such as databases and highly virtualized workloads. In 2015, larger enterprises will be leading with a “flash first” approach. A new emphasis will be on quantifying the benefits that an AFA solution can bring to the business. Every data center will be trying to justify moving to flash for all new tier 1 storage needs.
Prediction 2 – CEPH TM Will Move Beyond Pilot Implementations to Production in 2015 as Software Defined Storage Gains Acceptance
CEPH gained significant ground in 2014. RedHat announced the acquisition of Inktank (the company behind CEPH) April 30th, 2014. While obviously this brings CEPH and Inktank the credibility to gain acceptance in the enterprise, it’s been the adoption in the OpenStackTM community that has accelerated its adoption. We consistently hear CEPH and OpenStack used in same sentence in terms of new technologies being piloted together by enterprise customers. While this may seem unsurprising given RedHat’s support for both, it’s the broad support from Ubuntu TM and MirantisTM that is proving just as influential.
While CEPH provides file, object and block interfaces, it is at its core object storage. Both the CEPH FS file system and CEPH RADOS Block Device (RBD) are overlaid on the core object storage. The RADOS gateway also provides interfaces that are compatible to both Amazon’s S3TM and OpenStack’s SwiftTM. We have seen customers piloting CEPH with commodity hardware to deliver lower cost, and more flexible storage options this year. We believe this will be key to the adoption of CEPH and more broadly software defined storage overall in production in 2015. While providing flexibility customers should consider validating the performance needs of their storage workloads before committing to production usage in 2015.
Prediction 3 – Customers Will Continue to Leverage Cloud Storage for New Applications and Workloads in 2015
Cisco, in their Global Cloud Index for 2014, forecast that 78% of workloads will be delivered from a cloud data center. While this definition includes both public and private clouds (with the later accounting for 69% of those cloud workloads) it illustrates a continued fundamental shift to cloud based technologies – at least for non-mission critical applications. We consistently hear that today customers are looking to cloud storage for low cost alternatives for archiving, backup and document storage (think Box.net, Dropbox etc).
Cisco also forecasts that 53% of consumers will use cloud storage by 2018, enabling access from multiple devices. It will be needed to support new consumer facing mobile apps within the enterprise that will drive many of the new workloads to the cloud in 2015. While cost and flexibility have been at the heart of the move to cloud storage. As more diverse workloads enter the cloud performance will become paramount. Performance validation will be required to ensure the right trade-offs are made without sacrificing the customer experience.