gistfile1.txt

# Name agent components
a1.sources = r1
a1.sinks = k1 hb1
a1.channels = c1 c2

# Source Information
a1.sources.r1.type = avro
a1.sources.r1.bind = x.x.x.x #IP of web server or 0.0.0.0 to allow connection from all
a1.sources.r1.port = xx      # Port to listen on

# Channel information
a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000

a1.channels.c2.type = memory
a1.channels.c2.capacity = 1000

## Write to HDFS
a1.sinks.k1.type = hdfs
a1.sinks.k1.channel = c1
a1.sinks.k1.hdfs.path = /flume-data/%{log_type}/%d%m%Y
a1.sinks.k1.hdfs.useLocalTimeStamp = true
a1.sinks.k1.hdfs.fileType = DataStream
a1.sinks.k1.hdfs.writeFormat = Text
a1.sinks.k1.hdfs.rollSize = 0
a1.sinks.k1.hdfs.rollCount = 10000
a1.sinks.k1.hdfs.rollInterval = 600

# Write to Hbase
a1.sinks.hb1.type = org.apache.flume.sink.hbase.AsyncHBaseSink
a1.sinks.hb1.channel = c2
a1.sinks.hb1.table = my_table_name
a1.sinks.hb1.columnFamily = col_family_name
a1.sinks.hb1.column = col_name
a1.sinks.hb1.batchSize = 2
a1.sinks.hb1.serializer.incrementColumn = icol

# Bind sink to channel
a1.sources.r1.channels = c1 c2