Jonathan Hsieh Introduces Flume to the Apache Hadoop community. Flume is used to collect data, usually log files, and control the flow of this data from servers into a destination like HDFS.