Github user andrewor14 commented on a diff in the pull request:
https://github.com/apache/spark/pull/2161#discussion_r16799065
--- Diff:
core/src/main/scala/org/apache/spark/scheduler/cluster/mesos/CoarseMesosSchedulerBackend.scala
---
@@ -139,16 +145,16 @@ private[spark] class CoarseMesosSchedulerBackend(
if (uri == null) {
val runScript = new File(sparkHome,
"./bin/spark-class").getCanonicalPath
--- End diff --
As discussed offline, we can't actually do this because the Mesos working
directory is not the Spark directory. In this mode, we need to `rsync` the
driver's Spark directory to the slaves, so the spark home should be the same on
both the driver and the executors for most use cases.
However, if the executors and the driver don't share the same spark home,
the only way to configure the executor side's Spark home is through
`spark.home`. This is not super intuitive and I will submit a separate PR (that
shouldn't block RC1) to fix this.
---
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]
