我们收到了来自客户端的一个错误，关于一个oozie作业由于outofmemory问题而失败。oozie作业有三到四个动作，其中一个动作是Hive动作。
hive操作显然执行了一个join，intern执行一个完整的表扫描。由于客户端的维护活动，没有进行定期清除，数据积累了几天。这导致Hive行动额外扫描了几天。
下面是错误的堆栈跟踪：

2018-06-15 00:54:28,977 INFO org.apache.hadoop.hive.ql.io.HiveContextAwareRecordReader: Processing file hdfs://xxx:8020/data/csv/7342/2018-06-14/17/1/Network_xxx.dat
2018-06-15 00:54:29,005 INFO org.apache.hadoop.hive.ql.exec.MapOperator: Processing alias ntwk for file hdfs://xxx:8020/data/csv/7342/2018-06-14/17/1
2018-06-15 00:55:04,029 INFO org.apache.hadoop.hive.ql.exec.MapOperator: 7 finished. closing... 
2018-06-15 00:55:04,129 INFO org.apache.hadoop.hive.ql.exec.MapOperator: 7 forwarded 6672342 rows
2018-06-15 00:55:04,266 INFO org.apache.hadoop.hive.ql.exec.TableScanOperator: 3 finished. closing... 
2018-06-15 00:55:04,266 INFO org.apache.hadoop.hive.ql.exec.TableScanOperator: 3 forwarded 0 rows
2018-06-15 00:55:04,513 INFO org.apache.hadoop.hive.ql.exec.ReduceSinkOperator: 2 finished. closing... 
2018-06-15 00:55:04,538 INFO org.apache.hadoop.hive.ql.exec.ReduceSinkOperator: 2 forwarded 0 rows
2018-06-15 00:55:04,563 INFO org.apache.hadoop.hive.ql.exec.TableScanOperator: 3 Close done
2018-06-15 00:55:04,589 INFO org.apache.hadoop.hive.ql.exec.MapOperator: DESERIALIZE_ERRORS:0
2018-06-15 00:55:04,616 INFO org.apache.hadoop.hive.ql.exec.TableScanOperator: 1 finished. closing... 
2018-06-15 00:55:04,641 INFO org.apache.hadoop.hive.ql.exec.TableScanOperator: 1 forwarded 6672342 rows
2018-06-15 00:55:04,666 INFO org.apache.hadoop.hive.ql.exec.ReduceSinkOperator: 0 finished. closing... 
2018-06-15 00:55:04,691 INFO org.apache.hadoop.hive.ql.exec.ReduceSinkOperator: 0 forwarded 0 rows
2018-06-15 00:55:04,716 INFO org.apache.hadoop.hive.ql.exec.TableScanOperator: 1 Close done
2018-06-15 00:55:04,741 INFO org.apache.hadoop.hive.ql.exec.MapOperator: 7 Close done
2018-06-15 00:55:04,792 INFO ExecMapper: ExecMapper: processed 6672342 rows: used memory = 412446808
2018-06-15 00:55:10,316 INFO org.apache.hadoop.mapred.TaskLogsTruncater: Initializing logs' truncater with mapRetainSize=-1 and reduceRetainSize=-1
2018-06-15 00:55:10,852 FATAL org.apache.hadoop.mapred.Child: Error running child : java.lang.OutOfMemoryError: Java heap space
    at org.apache.hadoop.io.compress.DecompressorStream.<init>(DecompressorStream.java:50)
    at org.apache.hadoop.io.compress.BlockDecompressorStream.<init>(BlockDecompressorStream.java:50)
    at org.apache.hadoop.io.compress.SnappyCodec.createInputStream(SnappyCodec.java:173)
    at org.apache.hadoop.hive.ql.io.RCFile$Reader.nextKeyBuffer(RCFile.java:1447)
    at org.apache.hadoop.hive.ql.io.RCFile$Reader.next(RCFile.java:1602)
    at org.apache.hadoop.hive.ql.io.RCFileRecordReader.next(RCFileRecordReader.java:98)
    at org.apache.hadoop.hive.ql.io.RCFileRecordReader.next(RCFileRecordReader.java:85)
    at org.apache.hadoop.hive.ql.io.RCFileRecordReader.next(RCFileRecordReader.java:39)
    at org.apache.hadoop.hive.ql.io.HiveContextAwareRecordReader.doNext(HiveContextAwareRecordReader.java:274)
    at org.apache.hadoop.hive.ql.io.CombineHiveRecordReader.doNext(CombineHiveRecordReader.java:101)
    at org.apache.hadoop.hive.ql.io.CombineHiveRecordReader.doNext(CombineHiveRecordReader.java:41)
    at org.apache.hadoop.hive.ql.io.HiveContextAwareRecordReader.next(HiveContextAwareRecordReader.java:108)
    at org.apache.hadoop.hive.shims.HadoopShimsSecure$CombineFileRecordReader.doNextWithExceptionHandler(HadoopShimsSecure.java:329)
    at org.apache.hadoop.hive.shims.HadoopShimsSecure$CombineFileRecordReader.next(HadoopShimsSecure.java:247)
    at org.apache.hadoop.mapred.MapTask$TrackedRecordReader.moveToNext(MapTask.java:215)
    at org.apache.hadoop.mapred.MapTask$TrackedRecordReader.next(MapTask.java:200)
    at org.apache.hadoop.mapred.MapRunner.run(MapRunner.java:48)
    at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:417)
    at org.apache.hadoop.mapred.MapTask.run(MapTask.java:332)
    at org.apache.hadoop.mapred.Child$4.run(Child.java:268)
    at java.security.AccessController.doPrivileged(Native Method)
    at javax.security.auth.Subject.doAs(Subject.java:415)
    at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1408)
    at org.apache.hadoop.mapred.Child.main(Child.java:262)

我们收到了客户关于数据量的统计数据。看起来他们每天能收到300mb的数据。失败的配置单元查询已经处理了三天的数据，因此将近1gb的数据。三天共有3000万条记录
我们试图在我们的实验室中重现同样的错误，我们加载了总计1亿的模拟数据（五天内总共5gb的数据），持续了五天，但Hive查询和它的后台作业仍然无缝运行。
不知道为什么使用相同的map jvm参数，我们无法得到outofmemory错误。请注意，我们没有客户数据转储。我们使用的是模拟数据。
尽管数据量增加了五倍，但我们没有面临与客户相同的问题的原因可能是什么？无法理解原因。
配置如下：

mapred.map.child.java.opts : -Xmx512M
mapred.job.reduce.memory.mb : -1
mapred.job.map.memory.mb : -1