Manoj Kumar

Hi @kelvin-qin, Looks like you are also hit by [Driver issue](https://github.com/oap-project/gazelle_plugin/issues/1096#issue-1358947979) If its possible could you please share plan ?

Hi @weiting-chen @PHILO-HE **[Analysis]** **spark.sql.join.preferSortMergeJoin** When we enable above setting, it will favour shuffled hash join. So before that it try to coalesce shuffled batches. And it seems issue is...

Q6 also impacted looks multiple queries impacted by this coalesce issue 

@PHILO-HE Could you please validate my analysis ?

Hi @PHILO-HE, @weiting-chen , I get your point about the warning message. In my opinion there will be only 2 fix 1. Remove limit of 64k on batch size, but...

@PHILO-HE I am on it will update you. I more point could you please update on [driver side hash table build ](https://github.com/oap-project/gazelle_plugin/issues/1096) Because of driver side issue some querries getting...

@PHILO-HE query working fine on Q4 using main branch, I will do complete testing across multiple querries (TPCDS), once all validated. I will close this issue

@PHILO-HE When I compiled with recent main branch I get jar gazelle-plugin-1.5.0-SNAPSHOT-spark-3.2.1.jar. When I try to run that, am getting following error. Could you please help me in getting this...

@zhouyuan any idea why i am not able to run latest main branch on my system ? Getting above error

@PHILO-HE @zhouyuan wscg codegen file is empty