flume到hdfs和kafka

注意：需要将hadoop的相关jar包复制到flume下面

1. flume-conf.properties的配置

a1.sources = r1
a1.sinks = k1 sink-hdfs
a1.channels = c1 chn-hdfs

#tail -F 注意exec只适合测试，可以使用TAILDIR
a1.sources.r1.type = exec
a1.sources.r1.command = tail -F /home/abc/robotResume/jupiter/jupiter_http_log/logback.log
a1.sources.r1.inputCharset = UTF-8
a1.sources.r1.interceptors = i1
a1.sources.r1.interceptors.i1.type = timestamp

a1.sinks.k1.type=org.apache.flume.sink.kafka.KafkaSink
a1.sinks.k1.kafka.topic=nginx-resume
a1.sinks.k1.kafka.bootstrap.servers=localhost:9092
a1.sinks.k1.kafka.flumeBatchSize=100
a1.sinks.k1.kafka.producer.acks=1

a1.sinks.sink-hdfs.type=hdfs
a1.sinks.sink-hdfs.hdfs.fileType=DataStream
a1.sinks.sink-hdfs.hdfs.writeFormat=Text
a1.sinks.sink-hdfs.hdfs.path=hdfs://localhost:7000/flumeResume/data/%Y-%m-%d
a1.sinks.sink-hdfs.hdfs.rollInterval=0
a1.sinks.sink-hdfs.hdfs.rollSize=10240000
a1.sinks.sink-hdfs.hdfs.rollCount=0
#idleTimeout表示在此时间内没有写入则关闭连接并生成文件，0表示禁用此功能
a1.sinks.sink-hdfs.hdfs.idleTimeout=0
a1.sinks.sink-hdfs.hdfs.minBlockReplicas=1

a1.channels.c1.type = memory
a1.channels.c1.capacity = 1000
a1.channels.c1.transactionCapacity = 100
a1.channels.chn-hdfs.type=memory
a1.channels.chn-hdfs.capacity=1000
a1.channels.chn-hdfs.transactionCapacity=100

a1.sources.r1.channels = c1 chn-hdfs
a1.sinks.k1.channel = c1
a1.sinks.sink-hdfs.channel=chn-hdfs

 

2. sink hdfs文件滚动配置

hdfs.minBlockReplicas=1表示flume不需要去考虑hdfs的datanode复制备份。如果不设置1，其他roll设置会失效，flume会自动判断roll的条件。

3. sink hdfs定期配置

　　hdfs.round            默认false

　　hdfs.roundValue    值

　　hdfs.roundUnit      second, minute or hour

4. 按月生成文件夹，并按24小时生成文件

a1.sinks.sink-hdfs.type=hdfs
a1.sinks.sink-hdfs.hdfs.fileType=DataStream
a1.sinks.sink-hdfs.hdfs.writeFormat=Text
a1.sinks.sink-hdfs.hdfs.path=hdfs://localhost:7000/resumeLog/%Y-%m
a1.sinks.sink-hdfs.hdfs.rollInterval=86400
a1.sinks.sink-hdfs.hdfs.rollSize=0
a1.sinks.sink-hdfs.hdfs.rollCount=0 #event次数
a1.sinks.sink-hdfs.hdfs.idleTimeout=0 #在此时间内没有写入则关闭连接并生成文件，0禁用此功能
a1.sinks.sink-hdfs.hdfs.minBlockReplicas=1