Github user tgravescs commented on a diff in the pull request:
https://github.com/apache/spark/pull/634#discussion_r15173511
--- Diff:
yarn/common/src/main/scala/org/apache/spark/scheduler/cluster/YarnClientSchedulerBackend.scala
---
@@ -30,6 +30,11 @@ private[spark] class YarnClientSchedulerBackend(
extends CoarseGrainedSchedulerBackend(scheduler, sc.env.actorSystem)
with Logging {
+ if
(conf.getOption("spark.scheduler.minRegisteredExecutorsRatio").isEmpty) {
+ minRegisteredRatio = 0.8
--- End diff --
No real reason other then it might take longer to get 100%. Its just kind
of a number we choose to hopefully give the user a good experience without
having to wait to long if the cluster is really busy. The user can change it
if they want.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
