Plexistor said it has released results of benchmarking that showed performance improvements of MongoDB 3.2 with its new Software Defined Memory (SDM) architecture.
Plexistor´s converged SDM solution increases MongoDB performance by 450 percent, without sacrificing data persistency, in internal testing.
Open-source MongoDB databases offer features for enterprise-scale applications like analytics, serving content, and Internet of Things. MongoDB has been installed millions of times in hundreds of organizations around the world, including more than a third of the Fortune 100.
MongoDB achieves its best performance when the complete working data set fits in memory. However, with typical enterprise data sets reaching terabytes or petabytes, exceeding physical memory capacity, working sets are “sharded” or broken apart and distributed on multiple machines to maintain appropriate speed. Sharding is complex, incurs added costs, increases risk of data loss, and does not scale well.
With Plexistor´s SDM platform, MongoDB applications can achieve near-memory speed without expensive server hardware. Tests performed on a mid-range Intel CPU resulted in over 150,000 transactions per second while maintain transaction safety.
Plexistor was founded in Israel in 2013 and is headquartered in Mountain View, with its R&D in Herzliya, Israel.