Github user JoshRosen commented on a diff in the pull request:
https://github.com/apache/spark/pull/15986#discussion_r89256608
--- Diff:
core/src/main/scala/org/apache/spark/scheduler/TaskSchedulerImpl.scala ---
@@ -339,7 +341,7 @@ private[spark] class TaskSchedulerImpl(
// We lost this entire executor, so remember that it's gone
val execId = taskIdToExecutorId(tid)
- if (executorIdToTaskCount.contains(execId)) {
+ if (executorIdToRunningTaskIds.contains(execId)) {
reason = Some(
SlaveLost(s"Task $tid was lost, so marking the executor as
lost as well."))
removeExecutor(execId, reason.get)
--- End diff --
More archaeology:
It looks like `TaskState.LOST` was introduced in Spark 0.6.0 as part of a
refactoring to make the cluster scheduler pluggable:
https://github.com/apache/spark/commit/e72afdb817bcc8388aeb8b8d31628fd5fd67acf1.
That commit is from July 2012.
At the time, standalone mode didn't even exist and the schedulers were
Mesos and local mode, and only Mesos fine-grained mode was supported. The only
way to get a `TaskState.LOST` state was to convert the Mesos task loss state to
it.
---
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]
