[Enhancement](SparkLoad): avoid BE OOM in push task

[spaces-X]

Proposed changes

close #15572

Problem summary

Release memory pool held by the parquet reader when the data has been flushed by rowset writter.

Some test results:
1.unoptimized version

peak memory usage percent: 25%

2. optimized version

peak memory usage percent: 5%

Checklist(Required)

1. Does it affect the original behavior:
   • Yes
   • No
   • I don't know
2. Has unit tests been added:
   • Yes
   • No
   • No Need
3. Has document been added or modified:
   • Yes
   • No
   • No Need
4. Does it need to update dependencies:
   • Yes
   • No
5. Are there any changes that cannot be rolled back:
   • Yes (If Yes, please explain WHY)
   • No

Further comments

If this is a relatively large or complex change, kick off the discussion at dev@doris.apache.org by explaining why you chose the solution you did and what alternatives you considered, etc...

[github-actions]

clang-tidy review says "All clean, LGTM! 👍"

[hello-stephen]

TeamCity pipeline, clickbench performance test result:
the sum of best hot time: 36.5 seconds
load time: 647 seconds
storage size: 17122387725 Bytes
https://doris-community-test-1308700295.cos.ap-hongkong.myqcloud.com/tmp/20230104152925_clickbench_pr_73601.html

[morningman]

LGTM