[multistage][sorting] improve sorting efficiency

[walterddr]

currently we have a very basic implementation of the SortOperator and a very basic rule for cross-server sorting plan generator. It generates an sort exchange which connects a local sorting prior to exchange then follow by a remote sorting on a single machine this is originally designed for large quantity sorting and streaming responses. but none of those were implemented in SortOperator

Improvement

1. for the current SortOperator implementation, the sort-exchange can be simplified to a normal exchange, and the local sort prior to exchange can be removed.
   • [x] https://github.com/apache/pinot/pull/12237
   • [x] https://github.com/apache/pinot/pull/12251
2. for longer term we should support 2 strategies:
   • when data size is small, we should follow the 1st approach;
   • when data size is large, we should (a) create a local sort before the exchange, (either via sort-exchange, or via additional sort operator); (b) do a k-merge sort on the single-machine global sorting operator (c) stream the results back (which is possible with k-merge)

[hpvd]

@walterddr this looks great :-)

What do you think of how to process this issue best to keep a good overview:

• adding new issues for point 2 and enhance your tasklist?
• or closing this issue and opening a new one for Point 2?
• or..

having a tasklist for child issues is great. Having one we can also ad the new "parent issue label"...