org.apache.hadoop.mapred.maptask:开始刷新Map输出
2017-04-20 20:53:20,101 warn[main]org.apache.hadoop.mapred.yarnchild:运行子项的异常：java.lang.nullpointerexception位于org.apache.parquet.avro.avroschemaconverter.convertfield（avroschemaconverter）。java:294)在org.apache.parquet.avro.avroschemaconverter.convertfields（avroschemaconverter。java:204)在org.apache.parquet.avro.avroschemaconverter.convert（avroschemaconverter。java:198)在org.apache.parquet.avro.avroreadsupport.prepareforead（avroreadsupport。java:105)位于org.apache.parquet.hadoop.internalparquetrecordreader.initialize（internalparquetrecordreader）。java:174)在org.apache.parquet.hadoop.parquetrecordreader.initializeinternalreader（parquetrecordreader。java:192)在org.apache.parquet.hadoop.parquetrecordreader.initialize（parquetrecordreader。java:140)位于org.apache.hadoop.mapreduce.lib.input.delegatingrecordreader.initialize（delegatingrecordreader）。java:84)在org.apache.hadoop.mapred.maptask$newtrackingrecordreader.initialize（maptask。java:548)在org.apache.hadoop.mapred.maptask.runnewmapper（maptask。java:786)在org.apache.hadoop.mapred.maptask.run（maptask。java:341)在org.apache.hadoop.mapred.yarnchild$2.run（yarnchild。java:163)位于java.security.accesscontroller.doprivileged（本机方法）javax.security.auth.subject.doas（主题。java:415)在org.apache.hadoop.security.usergroupinformation.doas（usergroupinformation。java:1657)在org.apache.hadoop.mapred.yarnchild.main（yarnchild。java:158)
我尝试在mapper中读取filename.gz.paquet，使用相同的mapreduce作业，我可以读取filename.snappy.parquet文件。