 "GemStone Shares Grid Data Caching Expertise With Asian Market")
GemStone Systems, a provider of Enterprise Data Fabric (EDF), is on a mission to educate organisations about Grid Data Caching and help them increase CPU utilisation to nearly 100 percent.
At the Open Grid Forum, which took place recently in Singapore, Jags Ramnarayan, chief architect for GemStone Systems, led a group of IT professionals in an interactive, how-to session that explored how a scalable, flexible distributed main memory cache can operate in a grid environment to increase CPU utilisation to the highest levels.
“High availability of data is no longer enough; today, data has to be continuously available,” says Ramnarayan. “The recent volatility in the stock market is a prime example. It is essential for traders to have instant access to essential data so they can make quick, informed decisions; without it they would be paralysed”.
However, as the adoption of grid computing in the enterprise grows, several projects may be faced with a variety of challenges, according to Ramnarayan, most commonly, the data intensive nature of Grid applications can cause servers to become IO bound, sometimes reducing the average CPU utilisation to less than 50 percent.
Ramnarayan further said that applications should capitalise memory across the grid, as it is abundant. He recommended that IO cycles be kept to a minimum through main memory caching of operational data sets. Finally, he felt that horizontal partitioning of data and its behaviour can help achieve near infinite scale for Grid apps.