Can Trino use multiple writers to export a single CSV with an Order By query?

[ksg14]

Setup

• Trino v420
• Hive and delta connector
• Config variables:
  • scale-writers = false
  • task.writer-count = 1
  • writer-min-size = 1TB
  • distributed-sort = true
  • redistribute-writes = false
  • use-preferred-write-partitioning = false

Use-case

• Exporting a single CSV output (using CTAS query)

Question

1. Is use-preferred-write-partitioning = true and redistribute-writes = true only meant to sort each partition and not the final merged output?
2. With the current config i am able to get desired output but with a performance hit as i am using a single writer per worker per query. Can i leverage parallelism during the write phase and ensure sorting is achieved in final output?