You are viewing a plain text version of this content. The canonical link for it is here.

Posted to reviews@spark.apache.org by GitBox <gi...@apache.org> on 2020/11/03 13:49:04 UTC

[GitHub] [spark] gengliangwang edited a comment on pull request #30221: [SPARK-33314][SQL] Avoid dropping rows in Avro reader

gengliangwang edited a comment on pull request #30221:
URL: https://github.com/apache/spark/pull/30221#issuecomment-720397753


   So, if there is a scenario that calls `hasNextRow` multiple times, the code changes in https://github.com/apache/spark/pull/29145 also bring in perf regression:
   
   -  Before https://github.com/apache/spark/pull/29145, there is no deserialization in the method `hasNext` https://github.com/apache/spark/pull/29145/files#diff-a70c279fcf47a5f521902ddd83b7f4dd0f594c14b0f85240705faa83da79d859L104. 
   
   - After https://github.com/apache/spark/pull/29145, there is always deserialization if there is a next row: https://github.com/apache/spark/pull/29145/files#diff-22181c0e0050f9694efac388063535cf77e92a82dd962fec3f8507dfae45e52cR185
   
   I am sorry but shall we consider reverting https://github.com/apache/spark/pull/29145? CC @MaxGekk @cloud-fan 
   


----------------------------------------------------------------
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:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org