Toshiba Demonstrates Scale-Out Big Data Storage System

Even though it’s better known for enterprise and consumer laptop PCs and storage drives, Toshiba is investing in marketing entire new data center systems.

The venerable Irvine, Calif.-based company unveiled some new machines May 18 that are designed to address the hot categories of scale-out object storage, big data analytics, virtualization and active archives.

One of them is a multi-device storage solution that integrates Ethernet, NAND flash drives for low latency, large-capacity hard-disk drives (HDD) for higher throughput, 64-bit compute, and an open-source Linux platform for running next-generation software-defined storage applications on an industry-standard 3.5-inch form factor.

The other storage system Toshiba announced is an Ethernet-based HDD-only version, primarily optimized for the emerging shingled magnetic recording (SMR) media interface. This large-capacity system is optimized for archival and cold storage applications.

Scale-out storage is a type that uses a scaling methodology to create a dynamic storage environment that will support balanced data growth on an as-needed basis. Scale-out storage architecture uses a number of storage nodes—consisting of multiple low-cost computer servers and storage components—configured to create a storage pool or are configured to increase computing power to exceed a traditional storage array.

Using scale-out storage, the approach is to add storage nodes that work in tandem as users require additional storage resources. The scale-out architecture is designed to scale both capacity and performance.

As scale-out object storage deployments in enterprises increase, lower-cost systems have the potential to be part of mainstream performance as well as capacity-oriented business applications. Enabled with scale-out storage software such as Ceph, Toshiba’s high-performance and capacity-optimized object-storage drives are well-suited for enterprise primary storage, unstructured data, information governance, analytical data, and archival and cold storage.

Toshiba sees them as homogenous building blocks that enable enterprises to buy a single class of product and provision them differently for disparate and demanding workloads, thereby delivering a true software-defined storage infrastructure.

“When we were first designing Ceph 10 years ago, the key idea was that a loosely coordinated collection of smart devices may scale and perform better than a traditional array of disks. It is good to see that vision shared by leading component manufacturers like Toshiba and translated into a technology that can make its way into users’ hands,” said Sage Weil, head of the Ceph project and manager of software engineering and consulting engineer at Red Hat.

The multi-device demo is being showcased at booth P1 during OpenStack Summit 2015 in Vancouver, B.C., May 18 to 22. They are expected to be ready for market by late 2015.

For more information, go here.