As a journalist at iTnews:
Red Hat is launching a new platform for managing next-generation architecture in enterprise data centres and High Performance Computers (HPC).
Dubbed Enterprise Messaging, Realtime, Grid (MRG), Red Hat’s new offering is said to highlight a long-term industry trend towards utility computing.
The software integrates a standardised messaging platform and realtime capabilities with grid management technology, and could speed communications, promote interoperability and enable more flexible load and resource management.
According to Red Hat’s Global Product Manager Bryan Che, current e-mail and messaging offerings tend to be diverse, specialised, and lack interoperability.
In Sydney this week to meet with Red Hat’s local team and customers about Enterprise MRG, Che explained that a lack of interoperability could mean that organisations are faced with architectural challenges and are unable to reap the benefits of a holistic messaging ecosystem.
“If you look at the messaging space today, there’s no standard for messaging,” he said. “This means that if you buy one product from Vendor A, and another from Vendor B, they are not interoperable with each other.”
“[Enterprise MRG] is a pretty transformative stand for the industry,” he said. “We believe there to be some pretty transformative effects in how applications are written and how they operate with each other.”
To date, Red Hat’s messaging platform has been deployed by organisations such as Goldman Sachs and Credit Suisse, as well as JP Morgan, with whom the software was developed in collaboration.
The messaging platform is supported by a realtime component, which enables deterministic performance of a system using fine-grained control.
Red Hat’s realtime technology was developed in collaboration with the upstream Linux kernel community and has been optimised for use in the standard Red Hat Enterprise Linux environment.
The grid component of Enterprise MRG builds on the University of Wisconsin’s Condor Project, which was first developed in the 1980s and now is used in the Open Science Grid and IBM’s Blue Gene supercomputers.
The technology is expected to enable organisations to manage their resources and load with greater dynamic flexibility.
Similar to virtualisation, grid technology handles load using a shared pool of resources.
However, the grid technology further adds the ability to integrate traditional server resources with resources such as unused desktops and cloud computing resources.
“The virtualisation model is the first step towards the eventual goal of ‘I want to be much more efficient and how do I best manage my resources’,” Che explained.
“I think there is a lot of converging trends,” he said, describing Enterprise MRG and a rising demand for utility computing.
“Fundamentally, we’re going beyond HPC; what we’re looking to do [with Enterprise MRG] is provide you with the capability to take advantage of any company resource available to you,” he said.
Since it was publicly announced in December 2007, Enterprise MRG has been made available via limited release in North America and Europe.
Enterprise MRG will be available in Australia by the end of the year, when version 1.1 of the software is released globally.
