Github user yhuai commented on a diff in the pull request:
https://github.com/apache/spark/pull/8909#discussion_r41457829
--- Diff: sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveQl.scala ---
@@ -264,12 +262,16 @@ private[hive] object HiveQl extends Logging {
* Returns the HiveConf
*/
private[this] def hiveConf: HiveConf = {
- val ss = SessionState.get() // SessionState is lazy initialization, it
can be null here
+ var ss = SessionState.get()
+ // SessionState is lazy initialization, it can be null here
if (ss == null) {
- new HiveConf()
- } else {
- ss.getConf
+ val original = Thread.currentThread().getContextClassLoader
+ val conf = new HiveConf(classOf[SessionState])
+ conf.setClassLoader(original)
+ ss = new SessionState(conf)
+ SessionState.start(ss)
--- End diff --
Actually, I guess we are fine. When we create a `HiveConf`, its internal
classloader will be `Thread.currentThread().getContextClassLoader()`. So, we
will not change the context classloader. But, it is better to have explicitly
call `conf.setClassLoader` to set `original` as the classloader for this conf
to avoid any confusion. What do you think?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
