[BUG] REPL write query result failed when result size exceeds 10MB

[penghuo]

What is the bug?
REPL write query result failed when result size exceeds 10MB, no result write to query result index, no error message.

How can one reproduce the bug?

select * from table_1GB;

What is the expected behavior?

1. better error message is result is too large.

Do you have any additional context?

Job aborted due to stage failure: Serialized task 31:0 was 1350233713 bytes, which exceeds max allowed: spark.rpc.message.maxSize (134217728 bytes). Consider increasing spark.rpc.message.maxSize or using broadcast variables for large values.