apapi edited a comment on issue #27077: [SPARK-30408][SQL] Should not remove 
orderBy in sortBy clause in Optimizer
URL: https://github.com/apache/spark/pull/27077#issuecomment-571900240
 
 
   @maropu 
   My example in description is ambiguous, I updated the example.
   
   More details:
   For 
   ```val sortData = dataset.orderBy("b").sortWithinPartitions("c")```
   
   In Analyzed Logical Plan of sortData, there are 2 Sorts, 
   the first is is a local sort, which is the result of 
```sortWithinPartitions("c")```.
   The second is a global sort, which is the result of ```orderBy("b")```.
   
   But in the Optimized Logical plan, there is only one Sort, it is the first 
local sort, which is the result of ```sortWithinPartitions("c")```. The global 
sort (which is the result of ```orderBy("b")```) was removed by the Optimizer.
   
   In the code ```dataset``` was first repartitioned by ```orderBy()```, then 
content within each partition sorted by ```sortWithinPartitions()```.
   ```orderBy()``` may lead to repartition, but ```sortWithinPartitions()``` 
will not.
   So remove global sort lead to the lost of the repartition operation, which 
is not correct.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to