Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/20377#discussion_r163920410
--- Diff:
sql/hive/src/main/scala/org/apache/spark/sql/hive/client/IsolatedClientLoader.scala
---
@@ -251,10 +251,8 @@ private[hive] class IsolatedClientLoader(
/** The isolated client interface to Hive. */
private[hive] def createClient(): HiveClient = synchronized {
- val warehouseDir =
Option(hadoopConf.get(ConfVars.METASTOREWAREHOUSE.varname))
if (!isolationOn) {
- return new HiveClientImpl(version, warehouseDir, sparkConf,
hadoopConf, config,
- baseClassLoader, this)
+ return new HiveClientImpl(version, sparkConf, hadoopConf, config,
baseClassLoader, this)
--- End diff --
Let me ask the question: what exactly is the problem with the argument I
added? It solves the issue without having to write all this code.
If you really dislike the argument for some odd reason, you can get the
config by iterating over the `Iterable` in `HiveClientImpl`, making an
operation that is currently O(1) to be O(n).
But I really don't understand why you guys care about this argument so
much. There are *2* call sites to that constructor in the whole code base, both
in the same method in `IsolatedClientLoader.scala`.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
