Flume写入到HDFS

a1.sources = r1
a1.sinks = k1
a1.channels = c1


# 指定sources类型
# netcat
a1.sources.r1.type = netcat
a1.sources.r1.bind =master1
a1.sources.r1.port = 10000
a1.sources.r1.channels = c1 


# syslogtcp
# a1.sources.r1.type = syslogtcp
# a1.sources.r1.host =master1
# a1.sources.r1.port = 10000
# a1.sources.r1.channels = c1 


# spooldir
# a1.sources.r1.type = spooldir
# a1.sources.r1.spoolDir=/home/data
# a1.sources.r1.fileHeader = true 
# a1.sources.r1.channels = c1 


# 指定channels类型
# channels
a1.channels.c1.type = memory 
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100
 
 
# 指定sinks为打印
# a1.sinks.k1.type = logger
# a1.sinks.k1.channel = c1
# a1.sinks.k1.hdfs.fileType=DataStream 


# HDFS_sinks
a1.sinks.k1.channel = c1
a1.sinks.k1.type = hdfs
a1.sinks.k1.hdfs.path = hdfs://master1:8020/flume
a1.sinks.k1.hdfs.filePrefix = spool
a1.sinks.k1.hdfs.fileSuffix = .txt 
a1.sinks.k1.hdfs.fileType=DataStream 
# #HDFS上的文件达到100MB时生成一个文件
a1.sinks.k1.hdfs.rollSize = 100000000
# #不按照条数生成文件
a1.sinks.k1.hdfs.rollCount = 1000000
# #HDFS上的文件达到1分钟生成一个文件
a1.sinks.k1.hdfs.rollInterval = 0
# 时间戳是否应四舍五入
a1.sinks.k1.hdfs.round = true
# 四舍五入到此值的最高倍数（使用hdfs.roundUnit配置的单位），要小于当前时间
a1.sinks.k1.hdfs.roundValue = 1
# 时间戳四舍五入的单位 - 秒，分钟，小时
a1.sinks.k1.hdfs.roundUnit = minute
a1.sinks.k1.hdfs.useLocalTimeStamp = true

如有错误，欢迎私信纠正，谢谢支持！