MapR Makes Hadoop Run Faster

MapR speeds up analysis enabling even more creative uses of big data analytics. MapR’s patent pending Lockless Storage Services provides a distributed read/write layer that eliminates local file system limitations, Java issues and the overhead and restrictions associated with a single name node.

MapR significantly improves MapReduce throughput. The underlying storage services architecture enables highly optimized shuffle operations for further performance advantages. In addition to overall speed improvements, MapR also excels at random I/O throughput with faster performance than any other Hadoop distribution.

MapR capabilities include: Faster than other distributions with half the hardware Higher MapReduce and HBase throughput (2X – 5X) Optimized shuffle Higher random I/O (5X to 100X) Automatic and transparent compression reduces network overhead and disk footprint Lockless architecture scales linearly with number of cores and nodes