Public signup for this instance is disabled. Go to our Self serve sign up page to request an account.
We should allow data source to report partitioning and avoid shuffle at Spark side
[Github] Pull Request #20201 (cloud-fan)