HoldAn open-source, non-relational, distributed database modelled after Google's Bigtable. It's run on top of Hadoop and HDFS (Hadoop Distributed File System) to provide BigTable-like capabilities for Hadoop, enabling real-time read/write access to large datasets.