[HUDI-7822] Resolve the conflicts between mixed hdfs and local path in Flink tests

[hehuiyuan]

Change Logs

Resolve the conflicts for which may be mixed hdfs and local path together

org.apache.flink.table.catalog.exceptions.CatalogException: Catalog hudi path /var/folders/k1/65gcjk_92ws2bjh3ftpz33fc0000gp/T/junit845658121822910144 does not exist.

	at org.apache.hudi.table.catalog.HoodieCatalog.open(HoodieCatalog.java:122)
	at org.apache.hudi.table.catalog.TestHoodieCatalog.beforeEach(TestHoodieCatalog.java:167)

Impact

It can use hdfs and local path to run test case.

The tempdir is creatd by local path. But hoodiecatalog may use hdfs path if the hadoop env exist

Risk level

none

Documentation Update

none

Contributor's checklist

• Read through contributor's guide
• Change Logs and Impact were stated clearly
• Adequate tests were added if applicable
• CI passed

[CTTY]

This looks fine. But just being curious here, why do we need to unify how we get hadoopConf? Is there any bigger refactoring/feature on the way?

[hehuiyuan]

This looks fine. But just being curious here, why do we need to unify how we get hadoopConf? Is there any bigger refactoring/feature on the way?

Hi , it only use the mehtod HadoopConfigurations.getHadoopConf to create Filesystem which can keep consistency

[hudi-bot]

CI report:

• 3d87799 Azure: CANCELED

Bot commands

@hudi-bot supports the following commands:

• @hudi-bot run azure re-run the last Azure build

[yihua]

Azure CI is green.