holdenk commented on a change in pull request #26440:
[WIP][SPARK-20628][CORE][K8S] Start to improve Spark decommissioning &
preemption support
URL: https://github.com/apache/spark/pull/26440#discussion_r362297926
##########
File path:
core/src/main/scala/org/apache/spark/scheduler/cluster/CoarseGrainedSchedulerBackend.scala
##########
@@ -529,6 +557,16 @@ class CoarseGrainedSchedulerBackend(scheduler:
TaskSchedulerImpl, val rpcEnv: Rp
logError(t.getMessage, t))(ThreadUtils.sameThread)
}
+ /**
+ * Called by subclasses when notified of a decommissioning worker.
+ */
+ private[spark] def decommissionExecutor(executorId: String): Unit = {
+ // Only log the failure since we don't care about the result.
Review comment:
I think this was previously to avoid the case where the executor marked its
self as decomissioned and stopped scheduling jobs before the driver knew to
stop trying to schedule jobs on it, but we've changed that behaviour so it is
no longer needed to do the expensive request.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
