Open source software solution provider Red Hat today announced that it has added erasure coding and cache tiering to Inktank Ceph Enterprise, the open software-defined storage clustering software that Red Hat acquired in May with its $175 million acquisition of startup Inktank. Inktank was launched in 2012 by the developers behind the Ceph software-defined storage component of OpenStack in order to commercialize the project.

The new erasure coding and cache tiering features in Inktank Ceph Enterprise 1.2 work together to allow for “more efficient use of raw capacity,” Ross Turk, Director of Product Marketing of Storage and Big Data at Red Hat, told SiliconANGLE.

Turk said that erasure coding and cache tiering work together to allow users to achieve “a blend of economy and performance” that is right for them. “Before these features were available, Ceph could be used for archival purposes,” Turk said. “However, these features allow it to be done with increased flexibility.”

The company said erasure coding makes Inktank Ceph Enterprise 1.2 suitable for archive and cold storage use by reducing the cost per gigabyte. “Erasure-coded storage back-ends allow users to store data with less raw cluster capacity, still maintaining its durability,” Turk explained. “Previously, durability in Inktank Ceph Enterprise was accomplished with replication: each object placed in the cluster was replicated ‘N’ number of times (with ‘N’ being configurable, typically three.) Erasure-coded back-ends are able to achieve similar durability with lower storage overhead.”

Ross Turk, Director of Product Marketing, Storage and Big Data at Red Hat

Meanwhile, cache tiering enables “hot” data to be moved onto high-performance media when it becomes active and “cold” data to be moved onto low-performance media when it is no longer active. Explaining a use case, Turk used Facebook as an example. (Note: Facebook is not a customer of Red Hat’s). “Facebook stores an amazing number of photos, only some of which are regularly accessed,” Turk said. “Those photos need to be stored in an economical way. 

“Facebook could use an erasure-coded storage back-end, which stores objects densely while ensuring that they stay available even if nodes fail, and add a cache pool comprised of flash media (or something else fast). Photos that are regularly accessed would be automatically served from the cache pool, and would expire from it when they were no longer in demand to make room for others.”

.

Calamari 1.2 also updated

.

Red Hat has also updated Calamari, the on-premise management and monitoring platform within Inktank Ceph Enterprise 1.2. The new graphical monitoring and diagnostic capabilities in Calamari 1.2 help users in updating the Ceph management platform to support management functions.

According to Red Hat, these updates will enable an administrator to manage the core functionality of the Reliable Autonomic Distributed Object Store (RADOS) storage cluster, including the ability to manage individual storage devices and pool policies. “Calamari in 1.0 and 1.1 was primarily a monitoring tool; in 1.2, Calamari now has features that allow users to adjust cluster, pool, and OSD [or object storage device] settings,” Turk said.

Inktank Ceph Enterprise 1.2 is supported on Red Hat Enterprise Linux 6 and 7, and on Ubuntu 12.04 and 14.04. “These platforms (along with CentOS) are the ones that are most in-demand among our customers and community,” said Turk. “We will provide future support as the market demands.”
.

Photo credit: Wander Boessenkool via photopin cc

Photo of Ross Turk courtesy of Red Hat