feat: serialize table operations by lock rather than queue

[ShiKaiWi]

Which issue does this PR close?

Prepare for #800

Rationale for this change

Currently, the operations on tables are serialized by the worker queue, which has a serious disadvantage -- no work steal, which may leading to hungry of some tables' operations because of one big table.

This PR will replace the worker queue with a simple lock for every table.

What changes are included in this PR?

• Replace the worker queue with lock for every table;

Are there any user-facing changes?

Remove the configs for the write workers.

How does this change test

Existing tests.

[codecov-commenter]

Codecov Report

Merging #814 (9fa1b8b) into main (a8ed20e) will decrease coverage by 0.33%.
The diff coverage is 79.79%.

❗ Current head 9fa1b8b differs from pull request most recent head 67543cc. Consider uploading reports for the commit 67543cc to get more accurate results

@@            Coverage Diff             @@
##             main     #814      +/-   ##
==========================================
- Coverage   68.02%   67.69%   -0.33%     
==========================================
  Files         300      300              
  Lines       47229    46510     -719     
==========================================
- Hits        32129    31487     -642     
+ Misses      15100    15023      -77     

Impacted Files	Coverage Δ
analytic_engine/src/instance/close.rs	0.00% <0.00%> (ø)
analytic_engine/src/instance/create.rs	90.90% <ø> (-1.65%)	⬇️
analytic_engine/src/manifest/details.rs	95.62% <ø> (ø)
common_util/src/runtime/mod.rs	96.29% <ø> (ø)
components/skiplist/src/list.rs	98.87% <ø> (-0.46%)	⬇️
server/src/handlers/prom.rs	68.44% <ø> (ø)
server/src/handlers/query.rs	0.00% <ø> (ø)
table_engine/src/engine.rs	53.52% <ø> (ø)
analytic_engine/src/compaction/scheduler.rs	72.60% <14.28%> (-1.81%)	⬇️
analytic_engine/src/instance/engine.rs	61.15% <50.00%> (-2.88%)	⬇️
... and 14 more

... and 1 file with indirect coverage changes

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[jiacai2050]

LGTM