4b4f8e2a25
### What changes were proposed in this pull request? This is a follow-up of https://github.com/apache/spark/pull/31671 https://github.com/apache/spark/pull/31671 has an unexpected behavior change that it uses a different Hadoop conf (`sparkContext.hadoopConfiguration`) to instantiate `FileSystem`, which is used to qualify the warehouse path. Before https://github.com/apache/spark/pull/31671 , the Hadoop conf to instantiate `FileSystem` is `session.sessionState.newHadoopConf()`. More specifically, `session.sessionState.newHadoopConf()` has more conf entries: 1. it includes configs from `SharedState.initialConfigs` 2. in includes configs from `sparkContext.conf` This PR updates `SharedState` to use the final Hadoop conf to instantiate `FileSystem`. ### Why are the changes needed? fix behavior change ### Does this PR introduce _any_ user-facing change? yes, the behavior will be the same before https://github.com/apache/spark/pull/31671 ### How was this patch tested? manually check the log of `FileSystem` and verify the passed in configs. Closes #31868 from cloud-fan/followup. Lead-authored-by: Wenchen Fan <wenchen@databricks.com> Co-authored-by: Wenchen Fan <cloud0fan@gmail.com> Signed-off-by: Wenchen Fan <wenchen@databricks.com> |
||
---|---|---|
.. | ||
src | ||
pom.xml |