Apache Accumulo is a key/value store based on the design of Google's BigTable. Accumulo stores its data in Apache Hadoop's HDFS and uses Apache Zookeeper for consensus. While many users interact directly with Accumulo, several open source projects use Accumulo as their underlying store.
MIT Lincoln Lab has documented ingest rates of 400k records per second per node. Accumulo also has a clear edge in scalability being able to support datasets in the 10s of PBs.