Ten Software-defined Storage Predictions for 2014

Top-Ten-Predictions---KateThis past year was full of exciting milestones for our EMC Advanced Software Division (ASD) team.  EMC ViPR made its debut atEMC World and became generally available on September 27th.   Some of the brightest minds in the business have worked to address the most pressing cloud, storage, and Big Data problems facing customers today. And now that ViPR is commercially available, we’ve doubled down to bring additional exciting product enhancements to market very shortly – but I’m getting ahead of myself.

There are still many challenges and market forces that will shape the future of storage, and I decided to ask some of the ASD experts what kind of disruptive, drastic, or just plaininteresting changes they foresee for future of IT.

Here’s what they had to say:

Prediction One: Enterprises and service providers will demand choice for delivering cloud storage
“As enterprises and service providers move forward with their journey to the cloud, there will be expectation from technology vendors to make cloud storage platforms available in various configurations. A small number will expect to build their own stack with cloud software from vendors tying in their own commodity hardware. A large number of customers will expect a turnkey appliance with all built-in cloud capabilities.”

-Vikram Bhambri, Senior Director, ASD Product Management
Prediction Two: Large IT environments will adopt commodity servers
“In 2014, larger IT environments will adopt standardized farms of commodity servers alongside traditional storage arrays. This trend will drive demand for storage delivered as a software application, ready to run on these farms of commodity servers. Rock-solid data protection and operational efficiency will separate the winners from the losers.”
-Manuvir Das, VP Engineering – ViPR Data Services

Prediction Three: Software-defined storage starts a race to the top
“For the last several years, pure-play cloud storage service providers have focused on the lowest $/GB and a race to the bottom that many enterprises and service providers feel they can’t win. In fact, some pure plays have found out the hard way that racing to the bottom is a losing game. Software-defined storage changes the game. Software-defined storage means multiple data services such as object, HDFS, key value, and more can be layered over heterogeneous storage – including commodity – and feature multiple form factors. Technology vendors will offer a platform that, for the first time, will allow enterprises and service providers to truly compete with pure-play public cloud services on price, value, and ease of use.”
-George Hamilton, Sr. Product Marketing Manager, ViPR Data Services

Prediction Four: Enterprises will want choice of storage systems to serve as Big Data repositories

“As Big Data analytics gains wider adoption, enterprises will want turnkey storage systems that can provide reliable long-term storage (with capabilities such as Geo-replication, Erasure coding, etc.) and also serve as their Big Data repositories. A persistent cloud-scale storage system without single points of failure will help prevent data movement into ephemeral HDFSstorage and allow in-place analytics. Big Data solutions that offer flexibility to use Hadoop compatible storage systems will gain popularity in the enterprise market.”
– Priya Lakshminarayanan, Director of Product Management, ViPR Data Services  

Prediction Five: Advances in data analytics will drive software-defined storage capabilities
“As enterprise and service providers deploy solutions that allow them to analyze the large data sets they have collected, new and interesting software-defined storage use cases will become viable. Capabilities such as predictive data migration, dynamic data resiliency, and fluid data locality will enable these industries to offer services and functionality to their customers that are impossible on traditional storage platforms. These capabilities will provide customers new choices for data management and allow infinite flexibility in how they consume storage.”
Matt Mendonza, Sr. Product Manager, Storage Resource Management Suite

Prediction Six: The software-defined storage stack will continue to evolve
“Today there are clearly separate lifecycles for software defined compute, networking, and storage. 2014 will bring further refinement in the concept of software defined data access, separate from the storage technology, including the ability to view the same storage via multiple data access paradigms.  This trend arguably began with the emergence of iSCSI years ago, as storage managed via a filesystem could now be viewed as a block device, and continued with the advent of thin provisioned systems, as filesystems were used to manage on demand, sparsely allocated block devices.  In 2014, expect to see a further blurring of lines between data access methods, as object storage systems allow data ingested via an object protocol (S3, Swift) to be analyzed via Hadoop/HDFS or post-processed via file-based tools (NFS, CIFS), with the results available either as objects or in structured, tabular storage (Key-Value stores or no-SQL DBs), all with in place processing of the storage with no data duplication or data movement.”
-Mark O’Connell, ViPR Data Services Architect, ASD

Prediction Seven: Enterprises will realize growing efficiencies with software-defined storage
“The concept of software-defined storage gets a lot of attention for the benefits it brings in delivering storage more quickly and efficiently. Centralizing and automating storage management across multiple storage types is only the beginning, however. By abstracting data management and presentation from the underlying physical infrastructure, the cloud model is truly delivered and new operations are now possible on data in-place. Technology vendors already offer capabilities such as storing objects on file-based storage and will soon introduce Hadoop on existing storage. With the ability to analyze data stores from a central point made possible by this new platform, enterprises and service providers will soon be able to integrate data analytic applications to get more intelligence out of data on existing physical storage without the need to move data over costly bandwidth to dedicated appliances.”
-Mark Prahl, Principal Product Marketing Manager, ViPR Controller

Prediction Eight: Third platform growth will require PaaS solutions that do it all
“In 2013, IDC predicts worldwide IT spending will exceed $2.1 trillion, up 5.7% from 2012. This change is being driven by, as IDC calls it, the 3rd platform. Cloud, mobile, Big Data, and social drive the industry growth, with cloud spending, in particular, increasing by 25% to $100B in 2014. While many vendors are delivering cloud capabilities and storage automation, cloud solutions need to go beyond IT on-demand. To win market share, PaaS, IaaS (Infrastructure-as-a-Service) and STaaS (Storage-as-a-Service) solutions will need to easily support de-facto APIs (like S3 and OpenStack Swift), and provide easy ways to harness data analytics without additional significant investments in new hardware and preferably no additional headcount.”
-Christopher Ratcliffe, VP of Marketing, ASD

Prediction Nine: The Internet of Things is exploding, and we’ll need better ways to analyze Big Data
“The infrastructure component of Big Data (servers, storage, networking, and cloud) is expected to grow exponentially, with the storage component growing at nearly 50%. Big Data is being driven by a measurement granularity revolution, often known as the Internet of Things). This revolution is driving exponential growth in unstructured content. Information becomes abundant, but data scientists, to make sense of this information, will continue to be scarce. More automation and analytics tools, as and new object-based and data storage approaches, will address Big Data scale and processing flexibility for structured/unstructured data growth. This flexibility can be interpreted in many ways. An example could be providing data analytics across an environment, exposing heterogeneous data stores to big data Hadoop applications—all without moving data to a separate file repository.”
-Harrison Roberts, Business Development, ASD

Prediction Ten: SDx Software-defined everything
“2014 will be the year of SDx— software-defined everything —but customers will move with caution. Enterprises and service providers will force clear definitions as they embrace software-defined technologies.  IT will be disaggregated as compute, network, and storage move independently into their respective software-defined categories.”  Read the whole prediction here.
-Amitabh Srivastava, President, Advanced Software Division

We’d love to hear from you—what are your storage predictions? Do you agree with ours? Stay updated by subscribing to receive Rethink Storage blog notifications in the right hand column.

About the Author: Kate Greenough