Startup RNA Networks Virtualizes Memory Across Multiple Servers
The idea of managing pooled memory to manage traffic is gaining currency in other quarters.
Startup RNA Networks is offering a new form of virtualization -- creating a shared memory pool from multiple servers' random access memories and using the pool as a single, logical device.
It's not the first time server memories have been consolidated into a single unit. Clustering systems spread data across multiple server memories and keep it coherent with the database operations going on beneath the cluster. The cluster, in a sense, has a unified memory for data synchronization.
Oracle's Coherence product, part of its Fusion middleware, partitions frequently accessed data in database operations across multiple server memories, for high-speed use by Java applications.
RNA Networks performs a similar function by imposing data coherence across the memory pool.
But a memory pool for a cluster or an Oracle Real Application Cluster has a single function: to satisfy the data needs of individual nodes in the cluster. RNA Networks seeks to convert the memory pool across servers into its own virtual device, capable of executing functions that weren't possible before.
Take high-speed messaging, for example.
A trader on Wall Street frequently sets up or stages a series of large trades without executing them, anticipating market conditions. When he spots a favorable trading moment, he wants to order the execution of one of his trades as fast as possible, before circumstances change. RNA Networks came out of 18 months of stealth mode Feb. 2 to announce it has a messaging system that could speed up such trading.
With the transactions waiting to be triggered in server memories, RNAmessenger can speed execution because both it and the data are operating in memory and can move at a pace close to the speed of light. Instead of exchanging messages that require a processor to execute transaction logic, RNAmessenger sends pointers to data sets, indicating what has been ordered to be traded; an exchange of the pointers allows completion of the trade.
"We off-load the processing work from the CPU to the network fabric," said Andy Mallinger, VP of marketing for the Portland, Ore., startup, in an interview.
"We are working with a hedge fund firm that was previously able to execute 63 transactions a second on their system. They wanted 10,000 transactions per second. Speeding up the trade was more important than the price of the trade," he noted. RNAmessenger, combined with RNAcache for data synchronization, was able to exceed the 10,000-transactions-per-second requirement, he said.
Google in the Enterprise SurveyThere's no doubt Google has made headway into businesses: Just 28 percent discourage or ban use of its productivity products, and 69 percent cite Google Apps' good or excellent mobility. But progress could still stall: 59 percent of nonusers distrust the security of Google's cloud. Its data privacy is an open question, and 37 percent worry about integration.
CIOs Get Smart About BIIT’s tried for years to simplify business intelligence efforts. Have visual analysis tools and Hadoop and NoSQL databases helped? Respondents to our 2014 InformationWeek Analytics, Business Intelligence, and Information Management Survey have a mixed outlook.
Join us for a roundup of the top stories on InformationWeek.com for the week of December 14, 2014. Be here for the show and for the incredible Friday Afternoon Conversation that runs beside the program.