Github user tdas commented on a diff in the pull request:
https://github.com/apache/spark/pull/849#discussion_r12927323
--- Diff: core/src/main/scala/org/apache/spark/deploy/SparkSubmit.scala ---
@@ -228,7 +228,9 @@ object SparkSubmit {
if (isUserJar(args.primaryResource)) {
jars = jars ++ Seq(args.primaryResource)
}
- sysProps.put("spark.jars", jars.mkString(","))
+ if (jars.nonEmpty) {
+ sysProps.put("spark.jars", jars.mkString(","))
+ }
}
--- End diff --
This change ensures that spark.jars is not set to an empty string by
SparkSubmit, if not jars have been specified. This is different from previous
behavior where empty string was being passed on downstream, to YARN, etc. What
are the repercussions of this?
@pwendell
---
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.
---
