[GitHub] [spark] WangGuangxin opened a new pull request #35460: [SPARK-38160][SQL] Shuffle by rand could lead to incorrect answers when ShuffleFetchFailed happend

Wed, 09 Feb 2022 02:04:27 -0800 


WangGuangxin opened a new pull request #35460:
URL: https://github.com/apache/spark/pull/35460


   <!--
   Thanks for sending a pull request!  Here are some tips for you:
     1. If this is your first time, please read our contributor guidelines: 
https://spark.apache.org/contributing.html
     2. Ensure you have added or run the appropriate tests for your PR: 
https://spark.apache.org/developer-tools.html
     3. If the PR is unfinished, add '[WIP]' in your PR title, e.g., 
'[WIP][SPARK-XXXX] Your PR title ...'.
     4. Be sure to keep the PR description updated to reflect all changes.
     5. Please write your PR title to summarize what this PR proposes.
     6. If possible, provide a concise example to reproduce the issue for a 
faster review.
     7. If you want to add a new configuration, please read the guideline first 
for naming configurations in
        
'core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala'.
     8. If you want to add or modify an error type or message, please read the 
guideline first in
        'core/src/main/resources/error/README.md'.
   -->
   
   ### What changes were proposed in this pull request?
   When we do shuffle on indeterminate expressions such as `rand`, and 
ShuffleFetchFailed happend, we may get incorrent result since it only retries 
failed map tasks.
   
   To illustrate this, suppose we have a dataset with two columns `(range(1, 5) 
as a, rand() as b)`, we shuffle by `b` using two map tasks and two reduce tasks.
   <img width="630" alt="image" 
src="https://user-images.githubusercontent.com/1312321/153171110-13887a78-565f-4cdf-afda-bb9053ea7ef0.png";>
   
   **When there is a fetch failed and we need to rerun map task 2,  the 
generated partitions maybe different compared with last attempt, and finally we 
get a duplicate record with a = 4**
   <img width="610" alt="image" 
src="https://user-images.githubusercontent.com/1312321/153172262-bf5e2c81-8db6-4b01-8b70-8f284b54b09d.png";>
    
   This is very similary to the bug in Repartition+Shuffle, which is fixed by 
https://github.com/apache/spark/pull/22112. 
   This PR try to fix this by reuse current machenism.
   
   
   ### Why are the changes needed?
   Fix data inconsistent issue.
   
   
   ### Does this PR introduce _any_ user-facing change?
   No
   
   
   ### How was this patch tested?
   Added UT
   


-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]



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

Reply via email to