Dan's Take

Simplifying the Grid

Univa NAVOPs and large-scale grid computing.

• By Dan Kusnetzky
• 05/27/2016

Univa's Robert Lalonde, VP and General Manager for NAVOPS, came by to bring me up to date on what Univa's been up to since I last spoke with company representatives, and to introduce NAVOPS Command.

Who is Univa?
Univa styles itself the "leading innovator in workload management and containerization solutions," and has been offering Grid Engine which is a "workload optimization" tool that originally came from Sun Microsystems.

Parallel Processing
Univa's Grid Engine is a distributed resource management system that some would call a "parallel processing monitor." Grid Engine competes with a number of other products to help enterprises manage large-scale grid processing applications, including: IBM's LSF, tools from both SAP and Oracle, and a number of open source projects such as Beowulf, Open MPI and Parallelia.

The goal of this type of computing is to harness together a large, to absolutely huge, number of systems to execute pieces of a single application concurrently. This type of software is at the center of research, architectural engineering, mechanical engineering, digital content creation and both automotive and aeronautical engineering. It also is at the heart of weather forecasting and geological engineering for oil and gas companies.

Univa NAVOPS
Univa has pressed Grid Engine into service to help companies provision, monitor and manage containerized applications based upon Docker, Kubernetes and ATOMIC. Univa claims that its free download of this technology makes it possible to rapidly provision and deploy container infrastructure five times faster.

Dan's Take: Taking The Pain Out Of Going Parallel
If we stop to consider industry trends in the area of big data analysis and the ongoing use of parallel processing to address analytical needs of enterprises, most would consider the use of this approach to be a complex, difficult, expertise-heavy and time-consuming computer science project. Breaking down applications (usually called "decomposing" applications) into chunks that can execute simultaneously can be difficult. Setting up a grid can be even more challenging.

Univa, as well as its competitors, have been doing their best to put this approach to work simply and easily. If your organization is embarking on a big data journey or already is using parallel processing for research, analysis or engineering, getting to know Univa Grid Engine and Univa NAVOPS will certainly be worth the time.