By Ibby Rahmani – Product Marketing
Big data environments like Hadoop (Hortonworks, Cloudera), NoSQL (MongoDB, Cassandra), and Analytics (Cognos, Tableau) are transforming the way enterprises manage their ever-growing repositories of information. To maximize business value, companies are focused not only on gathering data, but also on extracting meaningful, actionable insights. But to rapidly search through the nonproductive, non-revenue-producing data in storage and analyze it to identify business trends and reveal new opportunities, they need the appropriate software and tools.
Initially, businesses implemented big data infrastructures based around servers, because of the cost and complexity of network support and storage. Starting with DAS (direct-attached-storage) to address their capacity needs, many IT organizations quickly realized they faced significant management and availability challenges.
Now businesses that have already started building big data environments are looking for options to meet their growing capacity needs, but in many cases are overwhelmed by storage options. Nimble Storage has taken a unique approach with the Adaptive Flash platform, leveraging flash just where it’s optimal in order to provide cost-efficient performance. Plus, Nimble’s architecture enables big data customers to scale server performance independent of capacity, and vice versa.
The other big advantage of using Nimble shared storage in big data environments is that it greatly simplifies data management. For example, a Hadoop customer would traditionally create three-way mirrors to optimize availability, but with Nimble Storage businesses can leverage a RAID-based approach, reducing their server capacity requirement by 33 percent. Furthermore, Nimble’s inline compression compliments Hadoop’s native columnar compression feature to further reduce required capacity. These businesses are not worried about high availability, thanks to Nimble’s proven reliability (so-called five nines, or 99.999 percent availability), which is significantly higher than what businesses can realistically expect from direct-attached-storage.
Nimble’s shared storage has benefits in data protection as well, by eliminating the need for separate primary storage, backup, and disaster recovery systems. Using efficient point-in-time snapshots, Nimble lets you easily and efficiently protect and restore data and create applications and backups in minutes.
Businesses looking for site-level protection leverage Nimble Storage’s built-in replication feature, which facilitates efficient copying (mirroring) to a secondary array. Nimble transfers only compressed data to a secondary site – without the need for additional hardware or software licenses. This means disaster recovery is affordable and easy to manage. In fact, Nimble Storage’s customer base replicates over 50 percent of hosted workloads, substantially higher than the industry average. Plus, businesses have been able to use Nimble for both their traditional workloads – such as Oracle, SQL, Exchange, and SharePoint – as well as their big data environments.
With benefits around cost, capacity efficiency, management, high availability, and data protection, it makes sense for businesses to look at shared storage not only to support their big data applications, but also their traditional applications requirements.