[SUPPORT] Flink-Hudi Unable to use Hudi metadata with S3

**Describe the problem you faced**

Hi,

I was creating a Flink SQL stream pipeline in AWS EMR to compact data into a Hudi COW table. Because of S3 slowdown errors that occasionally happened during Hudi writes, I tried to turn on the metadata table to eliminate S3 file listing but ran into the following exception saying S3 FS doesn't support atomic creation.

This issue seems particular related to Flink as I have another Spark/Hudi based batch pipeline running in the same type of EMR cluster, and Hudi metadata table functionality is working as expected with S3 FS.

Can you please help me with this issue?

Exception:
```
Caused by: org.apache.hudi.exception.HoodieLockException: Unsupported scheme :s3, since this fs can not support atomic creation
at org.apache.hudi.client.transaction.lock.FileSystemBasedLockProvider.(FileSystemBasedLockProvider.java:89) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method) ~[?:1.8.0_402]
at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62) ~[?:1.8.0_402]
at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45) ~[?:1.8.0_402]
at java.lang.reflect.Constructor.newInstance(Constructor.java:423) ~[?:1.8.0_402]
at org.apache.hudi.common.util.ReflectionUtils.loadClass(ReflectionUtils.java:73) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.hudi.common.util.ReflectionUtils.loadClass(ReflectionUtils.java:123) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.hudi.client.transaction.lock.LockManager.getLockProvider(LockManager.java:118) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.hudi.client.transaction.lock.LockManager.unlock(LockManager.java:109) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.hudi.client.HoodieFlinkTableServiceClient.initMetadataTable(HoodieFlinkTableServiceClient.java:216) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.hudi.client.HoodieFlinkWriteClient.initMetadataTable(HoodieFlinkWriteClient.java:318) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.hudi.sink.StreamWriteOperatorCoordinator.initMetadataTable(StreamWriteOperatorCoordinator.java:347) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.hudi.sink.StreamWriteOperatorCoordinator.start(StreamWriteOperatorCoordinator.java:192) ~[hudi-flink1.17-bundle-0.14.1.jar:0.14.1]
at org.apache.flink.runtime.operators.coordination.OperatorCoordinatorHolder.start(OperatorCoordinatorHolder.java:181) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.scheduler.DefaultOperatorCoordinatorHandler.startOperatorCoordinators(DefaultOperatorCoordinatorHandler.java:165) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.scheduler.DefaultOperatorCoordinatorHandler.startAllOperatorCoordinators(DefaultOperatorCoordinatorHandler.java:82) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.scheduler.SchedulerBase.startScheduling(SchedulerBase.java:618) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.jobmaster.JobMaster.startScheduling(JobMaster.java:1130) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.jobmaster.JobMaster.startJobExecution(JobMaster.java:1047) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.jobmaster.JobMaster.onStart(JobMaster.java:439) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.rpc.RpcEndpoint.internalCallOnStart(RpcEndpoint.java:198) ~[flink-dist-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.rpc.akka.AkkaRpcActor$StoppedState.lambda$start$0(AkkaRpcActor.java:622) ~[flink-rpc-akka_f5fd373f-2282-403d-a522-72b822a720aa.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.concurrent.akka.ClassLoadingUtils.runWithContextClassLoader(ClassLoadingUtils.java:68) ~[flink-rpc-akka_f5fd373f-2282-403d-a522-72b822a720aa.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.rpc.akka.AkkaRpcActor$StoppedState.start(AkkaRpcActor.java:621) ~[flink-rpc-akka_f5fd373f-2282-403d-a522-72b822a720aa.jar:1.17.1-amzn-1]
at org.apache.flink.runtime.rpc.akka.AkkaRpcActor.handleControlMessage(AkkaRpcActor.java:190) ~[flink-rpc-akka_f5fd373f-2282-403d-a522-72b822a720aa.jar:1.17.1-amzn-1]
at akka.japi.pf.UnitCaseStatement.apply(CaseStatements.scala:24) ~[?:?]
at akka.japi.pf.UnitCaseStatement.apply(CaseStatements.scala:20) ~[?:?]
at scala.PartialFunction.applyOrElse(PartialFunction.scala:127) ~[flink-scala_2.12-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at scala.PartialFunction.applyOrElse$(PartialFunction.scala:126) ~[flink-scala_2.12-1.17.1-amzn-1.jar:1.17.1-amzn-1]
at akka.japi.pf.UnitCaseStatement.applyOrElse(CaseStatements.scala:20) ~[?:?]
at scala.PartialFunction$OrElse.applyOrElse(PartialFunction.scala:175) ~[flink-scala_2.12-1.17.1-amzn-1.jar:1.17.1-amzn-1]
```

**To Reproduce**

Flink SQL / Hudi configs

```
'path'='...',
'connector' = 'hudi',
'table.type' = 'COPY_ON_WRITE',
'precombine.field' = 'integ_key',
'write.precombine' = 'true',
'write.bucket_assign.tasks' = '10',
'write.task.max.size' = '2014',
'write.operation' = 'upsert',
'hoodie.datasource.write.recordkey.field' = 'key',
'write.parquet.max.file.size' = '240',
'index.bootstrap.enabled' = 'false',
'write.index_bootstrap.tasks' = '200',
'metadata.enabled' = 'true'
```

**Environment Description**
* Hudi version : 0.14.1

* EMR: 6.15.0

* Hive version : 3.1.3

* Hadoop version : 3.3.6

* Storage (HDFS/S3/GCS..) : S3

* Flink: 1.17.1

* Running on Docker? (yes/no) : No

**Additional context**

Add any other context about the problem here.

**Stacktrace**

```Add the stacktrace of the error.```

Start with FileSystemBasedLockProvider.java:89 and follow the metadata initialization path through LockManager.java, HoodieFlinkTableServiceClient.java, and StreamWriteOperatorCoordinator.java. Reproduce the supplied Flink SQL configuration on Hudi 0.14.1 with S3, then determine what change is needed for metadata initialization to complete without the atomic-creation exception.

Written by the indexing model from the issue text.