org.apache.flume.EventDeliveryException：java.lang.RuntimeException： java.lang.ClassNotFoundException：Class 未找到org.apache.hadoop.fs.s3native.NativeS3FileSystem
我已经在core-site.xml中添加了。
<property>
<name>fs.s3n.impl</name>
<value>org.apache.hadoop.fs.s3native.NativeS3FileSystem</value>
<description>The FileSystem for s3n: (Native S3) uris.</description>
</property>
bin / hadoop classpath 的/ opt / hadoop的/ hadoop的在/ etc / hadoop的：/选择/的hadoop / hadoop的/共享/ hadoop的/普通/ LIB / 中：/ opt / hadoop的/ hadoop的/共享/ hadoop的/普通/ ：/选择/ hadoop的/ hadoop的/共享/ hadoop的/ HDFS中：/ opt / hadoop的/ hadoop的/共享/ hadoop的/ HDFS / LIB / 中：/ opt / hadoop的/ hadoop的/共享/ hadoop的/ HDFS / ：/选择/ hadoop的/ hadoop的/共享/ hadoop的/纱线/ LIB / 中：/ opt / hadoop的/ hadoop的/共享/ hadoop的/纱线/ 中：/ opt / hadoop的/ hadoop的/共享/ hadoop的/映射精简/ LIB / 中：/ opt / Hadoop的/ Hadoop的/股/的Hadoop / MapReduce的/ 的:: /选择/ Hadoop的/ Hadoop的/股/的Hadoop /工具/ lib中/的中：/ opt / Hadoop的/ Hadoop的//的contrib /能力调度器/ 的.jar
Hadoop版本是2.7.1
Flume版本是1.6.0
Flume Agent配置：
flume1.sources =
flume1.channels = kafka-channel-allMsgs kafka-channel-user-match-served-stream
flume1.sinks = s3-sink-user-match-served-stream s3-sink-allMsgs
flume1.channels.kafka-channel-allMsgs.type = org.apache.flume.channel.kafka.KafkaChannel
flume1.channels.kafka-channel-allMsgs.brokerList = 10.0.1.175:9092 , 10.0.1.229:9092
flume1.channels.kafka-channel-allMsgs.zookeeperConnect = 10.0.1.60:2181
flume1.channels.kafka-channel-allMsgs.topic = allMsgs
flume1.channels.kafka-channel-allMsgs.groupId = s3_flume_events
flume1.channels.kafka-channel-allMsgs.readSmallestOffset = false
flume1.channels.kafka-channel-allMsgs.parseAsFlumeEvent = false
flume1.channels.kafka-channel-user-match-served-stream.type = org.apache.flume.channel.kafka.KafkaChannel
flume1.channels.kafka-channel-user-match-served-stream.brokerList = 10.0.1.175:9092 , 10.0.1.229:9092
flume1.channels.kafka-channel-user-match-served-stream.zookeeperConnect = 10.0.1.60:2181
flume1.channels.kafka-channel-user-match-served-stream.topic = user_match_served_stream
flume1.channels.kafka-channel-user-match-served-stream.groupId = s3_flume_matched_served_stream
flume1.channels.kafka-channel-user-match-served-stream.readSmallestOffset = false
flume1.channels.kafka-channel-user-match-served-stream.parseAsFlumeEvent = false
flume1.sinks.s3-sink-user-match-served-stream.channel = kafka-channel-user-match-served-stream
flume1.sinks.s3-sink-user-match-served-stream.type = hdfs
flume1.sinks.s3-sink-user-match-served-stream.hdfs.filePrefix = user_match
flume1.sinks.s3-sink-user-match-served-stream.hdfs.useLocalTimeStamp = true
flume1.sinks.s3-sink-user-match-served-stream.hdfs.path=s3n://<aws>:<aws>@bucket/served_message/%y-%m/%y-%m-%d
flume1.sinks.s3-sink-user-match-served-stream.hdfs.batchSize = 1024
flume1.sinks.s3-sink-user-match-served-stream.hdfs.rollCount=1270000
flume1.sinks.s3-sink-user-match-served-stream.hdfs.rollInterval=1800
flume1.sinks.s3-sink-user-match-served-stream.hdfs.rollSize=133169152
flume1.sinks.s3-sink-user-match-served-stream.hdfs.codeC=bzip2
flume1.sinks.s3-sink-user-match-served-stream.hdfs.fileType=SequenceFile
flume1.sinks.s3-sink-allMsgs.channel = kafka-channel-allMsgs
flume1.sinks.s3-sink-allMsgs.type = hdfs
flume1.sinks.s3-sink-allMsgs.hdfs.filePrefix = allMsgs
flume1.sinks.s3-sink-allMsgs.hdfs.useLocalTimeStamp = true
flume1.sinks.s3-sink-allMsgs.hdfs.path=s3n://<aws>:<aws>@bucket/all_Msgs/%y-%m/%y-%m-%d
flume1.sinks.s3-sink-allMsgs.hdfs.batchSize = 1024
flume1.sinks.s3-sink-allMsgs.hdfs.rollCount=1270000
flume1.sinks.s3-sink-allMsgs.hdfs.rollInterval=1800
flume1.sinks.s3-sink-allMsgs.hdfs.rollSize=133169152
flume1.sinks.s3-sink-allMsgs.hdfs.codeC=bzip2
flume1.sinks.s3-sink-allMsgs.hdfs.fileType=SequenceFile