EMC Tackles 'Big Data'

EMC has introduced the latest version of its Isilon scale-out NAS appliance, which scales to up to 15.5 petabytes in a single file system while providing customers the ability to lock any part of the data for regulatory purposes.

The company also increased the efficiency and performance of its Atmos cloud storage arrays.

Both the new Isilon and Atmos products, which were unveiled this week at the EMC World conference being held in Las Vegas, are targeting customers facing issues with what EMC calls "big data," said Pat Gelsinger, president of information infrastructure products at EMC.

Gelsinger defined big data as data which scales to multipetabytes of capacity, which is created or collected and stored in real time, is collaborative in terms of customer use, and is self-provisioning.

A key to managing big data is scale out, which is the ability to increase the performance of the storage as nodes are added to handle increasing capacity, Gelsinger said.

Both the Isilon scale-out NAS appliances and the Atmos cloud storage appliances are targeted at applications which deal with big data, he said.

The new Isilon IQ 108NL, based on technology EMC got with its acquisition last year of Isilon, is the first storage system from any vendor to scale to over 15 petabytes in a single file system and volume.

The ability to expand capacity is as simple as plugging in a new node and pressing a couple of mouse buttons, at which time the capacity is configured and ready to use, Gelsinger said.

Gelsinger said the IQ 108NL scales far beyond competitive products, including storage gear from Hewlett-Packard which scales to a maximum of 64 TBs, NetApp at 100 TBs, and IBM at 2 PBs.

EMC also introduced SmartLock, a new feature for Isilon scale-out NAS storage that allows customers to set immutable protection levels for big data assets.

Sujal Patel, president of EMC's Isilon Storage division and the founder and former CEO of Isilon, said SmartLock provides WORM (write once, read many) capability for large file systems, giving customers the ability to specific areas of their data for a specified number of years from being deleted or modified.

"Enterprises can make sure a subset of their data isn't tampered with for the next three to five years," Patel said.

EMC's Atmos 2.0, the newest version of its cloud storage appliance, has been upgraded with five times its previous performance, giving it the ability to ingest around 500 million objects per day, Gelsinger said.

It is also now compatible with the Amazon S3 public storage cloud, allowing customers to move data between the Amazon cloud and clouds based on the Atmos 2.0. Gelsinger said this allows customers to do things like testing and developing applications using the Amazon cloud and then moving that cloud to the Atmos 2.0 for production purposes.

Also included with Atmos 2.0 is GeoParity software which stores parts of data objects across distributed storage environments instead of the objects as a whole to increase storage efficiency, the new GeoDrive software which EMC said provides fast transparent access to the cloud for Windows users and applications, and a new software development kit with support for a variety of applications including Apple iOS.

EMC introduced several other technologies to support big data.

Next: Other Support For Big Data

EMC plans to provide full open-source support for the Apache Hadoop project, which develops open-source software for reliable, scalable, distributed computing.

That support includes the eventual release of software, appliance, and eventually virtual appliance versions of the Hadoop technology in connection with technology it got with last year's Greenplum acquisition.

EMC also expanded its Documentum technology with the addition of new on-demand cloud provisioning capabilities and the ability to do complete audit trails for compliance purposes, Gelsinger said.

Also new is Project Lightning, a new PCIe Flash technology for use in servers which allow workloads to be migrated between compute-intensive and capacity-intensive storage environments.

The Isilon IQ 108NL hardware product and SmartLock data retention software application are already available. The IQ 108NL is list priced at $123,500 per node, while SmartLock lists starting at $1,950 per node.

Atmos 2.0 and GeoDrive are slated to ship in the second quarter. Interoperability with Amazon S3 is expected to be in beta in the second quarter.