我正在尝试了解简单的distinct
查询的配置单元查询计划,并且其中一个阶段的输出混乱程度很小。
我有一个只有两列的简单表,即id和value。只需4行,如下所述。
数据:
hive> select * from temp.test_distinct;
OK
1 100
2 100
3 100
4 150
计划
hive> explain select distinct value from temp.test_distinct;
OK
Plan not optimized by CBO.
Vertex dependency in root stage
Reducer 2 <- Map 1 (SIMPLE_EDGE)
Stage-0
Fetch Operator
limit:-1
Stage-1
Reducer 2
File Output Operator [FS_6]
compressed:false
Statistics:Num rows: 2 Data size: 10 Basic stats: COMPLETE Column stats: NONE
table:{"input format:":"org.apache.hadoop.mapred.TextInputFormat","output format:":"org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat","serde:":"org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe"}
Group By Operator [GBY_4]
| keys:KEY._col0 (type: string)
| outputColumnNames:["_col0"]
| Statistics:Num rows: 2 Data size: 10 Basic stats: COMPLETE Column stats: NONE
|<-Map 1 [SIMPLE_EDGE]
Reduce Output Operator [RS_3]
key expressions:_col0 (type: string)
Map-reduce partition columns:_col0 (type: string)
sort order:+
Statistics:Num rows: 4 Data size: 20 Basic stats: COMPLETE Column stats: NONE
Group By Operator [GBY_2]
keys:value (type: string)
outputColumnNames:["_col0"]
Statistics:Num rows: 4 Data size: 20 Basic stats: COMPLETE Column stats: NONE
Select Operator [SEL_1]
outputColumnNames:["value"]
Statistics:Num rows: 4 Data size: 20 Basic stats: COMPLETE Column stats: NONE
TableScan [TS_0]
alias:test_distinct
Statistics:Num rows: 4 Data size: 20 Basic stats: COMPLETE Column stats: NONE
Time taken: 0.181 seconds, Fetched: 35 row(s)
困惑:
TableScan,“选择运算符”和“按运算符分组”显示他们处理了 4行,这对我来说很有意义。但是在按操作员分组之后的下一阶段不应该仅获得 2行。作为分组依据将删除其他行。
在我的 DAG 中,我可以看到mapper的输出仅是两行,而不是四行,但是与Plan不匹配。
我可以看错吗?