optimize: acquire lock return fail-fast code in redis-pipeline mode.

[objcoding]

• I have registered the PR changes.

Ⅰ. Describe what this PR did

When the local transaction and the global transaction are enabled, the branch registration fails to acquire the global lock, the lock holder is in the second-stage rollback, and the branch registration fails to be retried quickly, because the retry with the local transaction does not release the database lock , resulting in a two-phase rollback wait. Therefore, if a global lock is found in the Rollbacking state, the fail-fast code is returned directly.

Ⅱ. Does this pull request fix one issue?

#4267

Ⅲ. Why don't you add test cases (unit test/integration test)?

Ⅳ. Describe how to verify it

Ⅴ. Special notes for reviews

[codecov-commenter]

Codecov Report

Merging #4277 (d63ceee) into develop (4f8e4b1) will increase coverage by 0.23%.
The diff coverage is 20.00%.

❗ Current head d63ceee differs from pull request most recent head 10e4de8. Consider uploading reports for the commit 10e4de8 to get more accurate results

@@              Coverage Diff              @@
##             develop    #4277      +/-   ##
=============================================
+ Coverage      48.55%   48.79%   +0.23%     
- Complexity      3791     3799       +8     
=============================================
  Files            724      724              
  Lines          24236    24211      -25     
  Branches        2970     2967       -3     
=============================================
+ Hits           11768    11813      +45     
+ Misses         11229    11158      -71     
- Partials        1239     1240       +1     

Impacted Files	Coverage Δ
...o/seata/server/storage/redis/lock/RedisLocker.java	37.56% <20.00%> (+0.69%)	⬆️
...er/src/main/java/io/seata/server/ServerRunner.java	50.00% <0.00%> (-7.70%)	⬇️
...n/src/main/java/io/seata/common/util/IdWorker.java	77.08% <0.00%> (-6.25%)	⬇️
...torage/file/store/FileTransactionStoreManager.java	56.12% <0.00%> (-0.65%)	⬇️
...o/seata/server/coordinator/DefaultCoordinator.java	48.38% <0.00%> (-0.54%)	⬇️
...in/java/io/seata/server/session/GlobalSession.java	79.31% <0.00%> (+0.43%)	⬆️
...io/seata/core/rpc/netty/AbstractNettyRemoting.java	14.28% <0.00%> (+1.29%)	⬆️
...e/properties/registry/RegistryNacosProperties.java	51.72% <0.00%> (+5.77%)	⬆️
...igure/properties/config/ConfigNacosProperties.java	75.00% <0.00%> (+12.50%)	⬆️
...ource/sql/struct/cache/AbstractTableMetaCache.java	89.65% <0.00%> (+17.24%)	⬆️
... and 2 more

[funky-eyes]

LGTM 补充下1.5.0.md

[funky-eyes]

按标题的意思这个pr应该是个optimize

[objcoding]

按标题的意思这个pr应该是个optimize

是的，原则上是pr3733的一个优化，因为pr3733并没有打包release版本，还处于develop分支。

[slievrly]

LGTM