Grid computing has made a crucial step towards mainstream enterprise adoption, the technology’s advocates believe. Once the preserve of academia and scientific research, grid technology has now been taken up in key vertical sectors such as the financial services industry – a leap that can make all the difference, grid advocates say.
Open Grid Forum chief executive Mark Linesch says. “What we are seeing now is the next stage of maturity.” As this week’s OGF20 event in Manchester draws to a close, Linesch cites financial services, the pharmaceutical and energy industries and the complex manufacturing sector including aerospace and car plants as the next big areas of development for grid computing.
“It’s very application driven. You justify the development of grids based on specific applications that are going to offer competitive advantage,” he says.
Financial services is a frontrunner in enterprise grid development, Linesch says, using the technology not just in areas such as risk modelling but also for data integration, while internet-based companies such as search giant Google and online auction house eBay “think of grid as part of their infrastructure” and almost take grid computing for granted.
“You can think of it as a unifying architecture for distributed computing at scale,” he says. That use of grids in enterprise architectures is what “hints at mainstream adoption”, Linesch argues. Leading IT departments are already consolidating hardware, using virtualisation, automating IT business processes and “moving to grid-like shared infrastructures”.
It is hard to predict when grid computing will be acknowledged as firmly in the mainstream, he says. “Evolution happens in fits and starts.”
But Linesch says grid offers the sort of resilience and reliability that enterprises need. “Inherent in the notion of grid architecture there’s not only the virtualisation of resources, but [the idea that] if I knock one node out of the grid there’s inherent failover capacity. The grid has natural resilience because it’s about utilising and aggregating a set of resources.”
Enterprises can use “intelligent” grid middleware to set policies for recovery and failover, he says. “The different middleware platforms out there for grid computing have these capabilities.”
Linesch says enterprises can look at grid adoption in the same way as virtualisation, as a next step in the chain of consolidation, automation and virtualisation. “It took IT probably five years to move virtualisation from the test and development environment to be thought of as ready for production environments.”
Once the concept of controlling virtual machines and resources has taken hold, then the conceptual leap to grid computing “makes more sense to people”, he says.
Linesch is keeping a sharp eye on the progress of grid computing in the financial services industry. Finance houses are already using shared file systems and virtualisation and are moving into grid, he notes. “The traditional banking and finance vertical has been a very aggressive user of IT – it’s to do with the pursuit of IT for competitive advantage. And traditionally, they have been the predictor of what more mainstream organisations will do.”
Grid computing’s new role in the financial sector is already becoming noticeable. In both London and New York, the City financial firms “are well on top of this”, Linesch says. He is confident that where the banking houses go first, others will soon follow.