CSE ICON now offers horizontally scalable storage systems, permitting big-data analysis at scale.
Leveraging MapR-FS, CSE-ICON provides storage solutions which grow to meet the most demanding of storage needs. Capable of storing both structured and unstructured data, MapR-FS provides a cluster based file system which permits concurrent reads and writes directly to disk, greatly out pacing the speed of traditional Hadoop file system storage.
In addition to speed, the MapR-FS eliminates traditional Namenode failover issues, by leveraging a distributed Container Location Database (CLDB) which supports ZooKeeper managed redundancy and automatic fail over, simplifying location services and increasing up time.
MapR-FS supports traditional Hadoop workloads in addition to processing that leverages the ability to write in-place, greatly speeding up Big Data jobs with optimized block sizes, direct-to-disk writes, no JVM garbage collection pauses, and lock contention prevention.