Legacy Models of Software Licensing Hindering Grid Computing Deployments
- By John K. Waters
- May 16, 2005
Enterprise licensing models are “throttling grid computing deployments,”
so software companies need to develop new models if grid is to continue to grow
in commercial markets.
That’s the conclusion of a report from The 451 Group. The 77-page report,
“Grid Computing–The Impact of Software Licensing,” is the
fourth published from the 451 Grid Adoption Research Service, an investigation
into user experiences and vendor strategies.
Conventional software licenses built around what the firm calls legacy concepts
of enterprise application use—the number of computers an application runs
on—are fiscally impractical for implementations of grid applications.
A grid could comprise hundreds of machines. Another complication: unlike clusters,
which comprise homogeneous environments, grids are designed to work across many
The report also found that alternative purchase models are appearing in the
marketplace, suggesting that a change is under way “that will have a cumulative
and disruptive impact on vendor licensing policies and practices.”
“Software licensing for grids must be seen within the context of other
dynamics,” the report advises. “The ability to proactively manage
the use of software licenses based on business objectives and in dynamic, virtual
environments is not a grid-only issue. Grids are an important inflection point
in this transformation, but the issue is a broader one for enterprise IT.”
As grid makes its way from traditional scientific and academic markets into
widespread commercial adoption, licensing is just one of the questions enterprise
IT managers are facing, says Mark Linesch, chair of the Global Grid Forum. Questions
about standards, security, interoperability, infrastructure and corporate politics
are cropping up as corporations begin grokking the grid.
“The whole industry—both the research community and the enterprise
IT folks—is talking about the next stages in the evolution of distributed
computing,” Linesch tells AppTrends. “We use a lot of terms—grids,
virtualization, SOAs, adaptive, on-demand, real-time—but what we’re
all really talking about is providing more abstraction and more flexibility
for collaboration across organizational domains.”
As a community-initiated forum of researchers and practitioners developing
grid computing standards and architectures, the GGF is all about promoting the
pervasive adoption of grid computing for both traditional research and high-performance
computing areas, and a growing number of enterprise customers. “Our community
wants us to figure out how to standardize this stuff and how to share best practices,”
The GGF has also found that current enterprise licensing schemes are cost prohibitive
when it comes to grid applications. The forum has begun establishing “communities
of interest”—research groups—to explore mass adoption in the
financial, pharmaceutical and telecommunications industries.
“What the vendors need to understand is that grid is about a lot more
today than just sharing compute cycles,” Linesch says. “Grid is
about storage and data and scarce instrumentation and collaboration. It’s
still about scientific and high-performance computing, but it’s also about
a manufacturer who wants to share data across different organizational boundaries
to collaborate on the design of some airplane seats. It’s about a pharmaceutical
company that wants to do drug discovery with three other startups. It’s
about companies that want to bring together disparate data sources so they can
orchestrate more effective business processes.”
For more information about the GGF, go to www.ggf.org.
For more information about The 451 Group and its grid computing report, go to:
John K. Waters is a freelance writer based in Silicon Valley. He can be reached
at [email protected].