[fix](move-memtable) fix delta writer close after closed

[kaijchen]

Proposed changes

Fix concurrency issue where some delta writer meets "close after closed" issue.
Because some sink may be closed before other sinks to be opened,
the reference count dropps to 0 and the pool will be closed when the load is not done.

To fix the problem, we should init delta writer pool in sink prepare.
Because there is a barrier between prepare and open.
All the sinks should have been prepared before any of them may close.

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...

[kaijchen]

run buildall

[github-actions]

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

[doris-robot]

TeamCity be ut coverage result:
Function Coverage: 37.37% (8475/22676)
Line Coverage: 29.67% (68524/230974)
Region Coverage: 28.24% (35587/126014)
Branch Coverage: 25.11% (18217/72556)
Coverage Report: http://coverage.selectdb-in.cc/coverage/c537473ca6489fcec8aa112b09790e6c40f35d5f_c537473ca6489fcec8aa112b09790e6c40f35d5f/report/index.html

[doris-robot]

(From new machine)TeamCity pipeline, clickbench performance test result:
the sum of best hot time: 46.31 seconds
stream load tsv: 554 seconds loaded 74807831229 Bytes, about 128 MB/s
stream load json: 20 seconds loaded 2358488459 Bytes, about 112 MB/s
stream load orc: 65 seconds loaded 1101869774 Bytes, about 16 MB/s
stream load parquet: 32 seconds loaded 861443392 Bytes, about 25 MB/s
insert into select: 28.8 seconds inserted 10000000 Rows, about 347K ops/s
storage size: 17162099268 Bytes

[dataroaring]

LGTM

[github-actions]

PR approved by at least one committer and no changes requested.

[github-actions]

PR approved by anyone and no changes requested.

[TangSiyang2001]

LGTM