更改 parquet table 的数组 <struct<>> 类型的列会导致配置单元出错

Question

我有一个

类型的字段

array<struct<id:string>>

在配置单元中创建的分区镶木地板 table 中。当我更改此列的类型以在结构部分添加字段时，在某个select中展开数组时出错。详情如下：

创建table 语句

CREATE EXTERNAL TABLE `test_table`(
  `my_array` array<struct<id:string>>)
PARTITIONED BY ( 
  `ymd` int)
ROW FORMAT SERDE 
  'org.apache.hadoop.hive.ql.io.parquet.serde.ParquetHiveSerDe' 
STORED AS INPUTFORMAT 
  'org.apache.hadoop.hive.ql.io.parquet.MapredParquetInputFormat' 
OUTPUTFORMAT 
  'org.apache.hadoop.hive.ql.io.parquet.MapredParquetOutputFormat'

使用数据创建分区

insert overwrite table test_table
PARTITION (ymd = 20170101)
select
   array(named_struct('id', 'id1')) as my_array

然后我在数组的结构中添加一个新字段

ALTER TABLE test_table 
CHANGE COLUMN my_array my_array array<struct<id:string, amount:int>>

这会更改 table 的元数据。我的期望是过去的数据仍然可读，'amount' 为空值。不幸的是，我运行陷入了一个我不明白的错误。为了更好地说明，让我先创建一个新分区：

insert overwrite table test_table
PARTITION (ymd = 20170102)
select
array(named_struct('id', 'id2', 'amount',2)) as my_array

现在，做一个

select * from test_table

产生了我预期的结果（HUE UI 的输出）：

但是，我想这样用侧视图来分解数组时就出错了：

select
    my_array
from
    test_table t
    lateral view explode (my_array) arry as a

此查询引发配置单元运行时间错误。日志的相关部分应该是本段之后的日志。当 selecting 'arry.a' 而不是 'my_array' 时，会发生非常相似的错误。令人惊讶的是，下面的查询运行很好，结果符合我的预期：

select
    ymd,
    a.id,
    a.amount
from
    test_table t
    lateral view explode (my_array) arry as a

在我看来这可能是一个错误。这是运行上面 select 导致错误时的一段日志。配置单元版本为 1.1.0-cdh5.8.0:

Error: java.lang.RuntimeException: 
org.apache.hadoop.hive.ql.metadata.HiveException: 
Hive Runtime Error while processing row {"my_array":[{"id":"id1","amount":null}],"ymd":20170101} at org.apache.hadoop.hive.ql.exec.mr.ExecMapper.map(ExecMapper.java:179) at org.apache.hadoop.mapred.MapRunner.run(MapRunner.java:54) at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:453) at org.apache.hadoop.mapred.MapTask.run(MapTask.java:343) at org.apache.hadoop.mapred.YarnChild.run(YarnChild.java:164) 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:1693) at org.apache.hadoop.mapred.YarnChild.main(YarnChild.java:158) 
Caused by: org.apache.hadoop.hive.ql.metadata.HiveException: Hive Runtime Error while processing row {"my_array":[{"id":"id1","amount":null}],"ymd":20170101} at org.apache.hadoop.hive.ql.exec.MapOperator.process(MapOperator.java:507) at org.apache.hadoop.hive.ql.exec.mr.ExecMapper.map(ExecMapper.java:170) ... 8 more 
Caused by: java.lang.UnsupportedOperationException: Cannot inspect java.util.ArrayList at org.apache.hadoop.hive.ql.io.parquet.serde.ArrayWritableObjectInspector.getStructFieldsDataAsList(ArrayWritableObjectInspector.java:172) at org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.serialize(LazySimpleSerDe.java:355) at org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.serialize(LazySimpleSerDe.java:319) at org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.serializeField(LazySimpleSerDe.java:258) at org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.doSerialize(LazySimpleSerDe.java:242) at org.apache.hadoop.hive.serde2.AbstractEncodingAwareSerDe.serialize(AbstractEncodingAwareSerDe.java:55) at org.apache.hadoop.hive.ql.exec.FileSinkOperator.processOp(FileSinkOperator.java:668) at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:815) at org.apache.hadoop.hive.ql.exec.SelectOperator.processOp(SelectOperator.java:84) at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:815) at org.apache.hadoop.hive.ql.exec.LateralViewJoinOperator.processOp(LateralViewJoinOperator.java:133) at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:815) at org.apache.hadoop.hive.ql.exec.UDTFOperator.forwardUDTFOutput(UDTFOperator.java:125) at org.apache.hadoop.hive.ql.udf.generic.UDTFCollector.collect(UDTFCollector.java:45) at org.apache.hadoop.hive.ql.udf.generic.GenericUDTF.forward(GenericUDTF.java:107) at org.apache.hadoop.hive.ql.udf.generic.GenericUDTFExplode.process(GenericUDTFExplode.java:94) at org.apache.hadoop.hive.ql.exec.UDTFOperator.processOp(UDTFOperator.java:108) at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:815) at org.apache.hadoop.hive.ql.exec.SelectOperator.processOp(SelectOperator.java:84) at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:815) at org.apache.hadoop.hive.ql.exec.LateralViewForwardOperator.processOp(LateralViewForwardOperator.java:37) at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:815) at org.apache.hadoop.hive.ql.exec.TableScanOperator.processOp(TableScanOperator.java:95) at org.apache.hadoop.hive.ql.exec.MapOperator$MapOpCtx.forward(MapOperator.java:157) at org.apache.hadoop.hive.ql.exec.MapOperator.process(MapOperator.java:497) ... 9 more

Answer 1

我们通过编写自定义 SerDe 解决了这个问题，只要在文件中找不到列，它就会为数据提供空值。

更改 parquet table 的数组 <struct<>> 类型的列会导致配置单元出错

altering column of type array<struct<>> of parquet table leads to error in hive

hive

parquet

alter-table