Setup HDFS in GitHub CI

To execute unit tests with Java 17, we can't use MiniDFSCluster and have to rely on external HDFS. 

Utils to use external HDFS:

https://github.com/apache/hudi/blob/2a0223933884cb044e7aa56f205cae926358a030/hudi-common/src/test/java/org/apache/hudi/common/testutils/HoodieTestUtils.java#L241 

Currently, we have docker image set up and those Java 17 tests would be executed there ([https://github.com/apache/hudi/blob/2a0223933884cb044e7aa56f205cae926358a030/.github/workflows/bot.yml#L209)]

A more ideal solution would be setting up HDFS in GitHub CI directly, for example: https://github.com/beyondstorage/setup-hdfs

## JIRA info

- Link: https://issues.apache.org/jira/browse/HUDI-6600
- Type: Improvement

Start with .github/workflows/bot.yml around line 209 and compare its Docker-based Java 17 test setup with the external HDFS helpers in hudi-common/src/test/java/org/apache/hudi/common/testutils/HoodieTestUtils.java around line 241. Review the setup-hdfs example linked in the issue and identify how HDFS can run directly in GitHub CI. Done means the Java 17 unit tests use CI-provided HDFS without relying on the existing Docker test environment.

Written by the indexing model from the issue text.