Ganglia, an open source scalable distributed monitoring system for high-performance computing systems, announced today the release of Ganglia 3.1. Replete with new metric features and a new modular interface, Ganglia 3.1 will bring improved performance and greater configuration options to organizations such as the University of California, Flickr and Stanford Linear Accelerator. Designed for large-scale monitoring environments, Ganglia is based on a hierarchical design targeted at federations of clusters, achieving very low per-node overhead and high concurrency, and has been ported to an extensive set of operating systems and processor architectures. Ganglia has been deployed on thousands of clusters worldwide and is able to monitor clusters as large as thousands of nodes, both as stand-alone system and as a component of a wide variety of software solutions. “We use Ganglia for both capacity planning and problem identification purposes,” said John Allspaw, Operations Engineering Manager at Flickr, a Yahoo! company and one of the world’s fastest growing online photo-sharing communities. "We use Ganglia to collect and record metrics on every host at Flickr, and it fits our architecture quite well. We have the ability to write new metrics quickly, and aggregate those across many clusters automatically.” “Ganglia’s wide-spread adoption and its use in varied environments created significant demand for the new capabilities included in Ganglia 3.1,” said Brad Nicholes, Ganglia developer and senior software engineer at Novell. “As an enterprise monitoring solution as well as an open source project, adding a higher degree of flexibility that would support a broader set of monitored metrics, seemed like the right thing to do to move the Ganglia project to the next level." Key new features of 3.1 include:

• Introduction of a modular interface for C and Python.
• All pre-existing metrics (CPU, network, disk, memory, etc.) have been converted into pluggable modules, which can be selected a la carte, allowing for a tailored monitoring environment.
• New pluggable metric modules such as multicpu, multidisk and tcp_conn status have been introduced in addition to the standard metrics, allowing for more accurate monitoring.
• The metric metadata has been expanded to include additional information such as metric groups, metric titles and descriptions, which can be displayed on the front-end view, or grouped and hidden as the user desires.
• Existing graphing capabilities have been modularized to facilitate creation of custom report graphs.
• Additional scaling capacity allows more accurate monitoring for systems greater then 4TB.

“Ganglia is an important part of GroundWork’s solution offerings, and one of our most successful community partnerships. The Ganglia Development Team has created a highly-scalable IT monitoring tool that’s allowed us to offer increased capacity and performance to our largest customers,” said David Lilly, founder and CEO of GroundWork Open Source. “GroundWork Open Source offers a Ganglia Integration Module that enables GroundWork Monitor to provide multiple role status views, dashboards, reports, notifications and configuration tools for data collected and displayed by Ganglia. Some of our largest deployments use Ganglia and GroundWork Monitor together in an integrated fashion to monitor over 10,000 servers in business critical environments.” Ganglia 3.1 will be available immediately and can be downloaded from its Web site. The project, as well as documentation, project details and more, can also be found at sourceforge.net/projects/ganglia/ and freshmeat.net/projects/ganglia/.