Red Hat Targets Petabyte Scale, Big Data With Storage Server 3

Red Hat on Thursday unveiled a major upgrade to its Red Hat Storage Server technology to give it petabyte scalability, improved data protection and easy integration into Red Hat Enterprise Linux deployments.

The new Red Hat Storage Server 3 is the second major release of the technology since Red Hat in late 2011 purchased Gluster, developer of the GlusterFS scale-out NAS file system on which it is based.

Red Hat Storage Server 3 has at its heart the Gluster version 3.6 open-source, software-defined technology, said Irshad Raihan, senior principal product marketing manager for storage and big data at Red Hat.

[Related: Red Hat's $175M Inktank Buy Expands Open Source Storage Reach With Ceph]

It was designed to help customers better manage the explosion in the amount of data being stored, the increasingly complex nature of enterprise storage and a need for increased control over data for regulatory and governance purposes, Raihan told CRN.

"Line of business administrators have become big buyers of IT," he said. "But today, IT is not flexible. There is a lot of proprietary storage. Open-source, software-defined storage is more flexible and provides the best efficiency."

Red Hat Storage Server 3 has five new features distinguishing it from earlier versions, Raihan said.

The first is a big jump in capacity to up to 19 petabytes per cluster. Red Hat Storage Server 3 can be configured with up to 128 nodes, each with up to 60 storage drives, up from a maximum of 64 nodes with 36 drives in the previous version.

It also includes built-in data protection via on-line snapshots to provide point-in-time copies for recovery and allows end users to recover files and volumes without administrator involvement, he said.

Also new is the ability to easily add the technology as a layer on top of existing Red Hat Enterprise Linux implementations via RPM, formerly known as Red Hat Package Manager.

The fourth is improved monitoring and logging of storage operations and issues using the Nagios open source monitoring service which plugs directly into the Red hat storage console.

Finally, Red Hat added a plug-in for the Hadoop File System (HDFS) to make it easier to use the storage system for big data operations, Raihan said.

NEXT: Red Hat Storage Server 3 For Hadoop, Other Workloads

"If a customer wants to run analytics on data in Red Hat storage today, they need to migrate the data to Hadoop and then back, which is expensive," Raihan said. "We are working with HortonWorks to develop the ability to run Hadoop analytics directly on data in Red Hat Storage Server. This works with a wide range of Hadoop analytic and management tools."

Red Hat also unveiled reference architectures aimed at helping customers take better advantage of Red Hat Storage Server 3.

For big data, Red Hat is working with IBM and Lenovo servers and with HortonWorks' Hadoop software. For cyber security analytics, Red Hat is partnering with Cisco for servers and Splunk for the software. The company is also working with Hewlett-Packard servers and the open source ownCloud file sync and share software to provide enterprise collaboration and file sharing capabilities, Raihan said.

Red Hat has come a long way with its storage offering since its original acquisition of Gluster, said Bradley Brodkin, president and CEO of HighVail Systems, a Toronto, Ont.-based solution provider and Red Hat channel partner.

Previous versions lacked a lot of the functionality customers required, Brodkin told CRN.

"With version 1, the original Gluster technology, it was great for building out a simple solution, but was very limited in terms of scalability, which required a lot of skill to do," he said. "While that meant services opportunities for partners, customers weren't interested in building out a solution they could get as an integrated solution elsewhere."

Version 2 was better, but not yet ready, Brodkin said. However, the issues seem to have been solved with version 3.

The integration with Red Hat Enterprise Linux is an important competitive edge.

"This takes Red Hat up to and maybe beyond what VMware is talking about with its storage capabilities," he said. "This brings storage much closer to the server. And using software to work with storage makes it easier to scale."

A couple of Red Hat solution providers told CRN on condition of anonymity that Red Hat Storage Server has had a rocky history. One channel partner said Gluster had great technology as a stand-alone open source storage developer, but customers were not excited about having to pay for support after the Red Hat acquisition.

However, the solution providers said, the latest version overcomes those issues.

PUBLISHED OCT. 2, 2014