Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/14710#discussion_r76893461
--- Diff:
core/src/main/scala/org/apache/spark/scheduler/cluster/CoarseGrainedSchedulerBackend.scala
---
@@ -532,39 +547,53 @@ class CoarseGrainedSchedulerBackend(scheduler:
TaskSchedulerImpl, val rpcEnv: Rp
final def killExecutors(
executorIds: Seq[String],
replace: Boolean,
- force: Boolean): Boolean = synchronized {
+ force: Boolean): Boolean = {
logInfo(s"Requesting to kill executor(s) ${executorIds.mkString(",
")}")
- val (knownExecutors, unknownExecutors) =
executorIds.partition(executorDataMap.contains)
- unknownExecutors.foreach { id =>
- logWarning(s"Executor to kill $id does not exist!")
- }
- // If an executor is already pending to be removed, do not kill it
again (SPARK-9795)
- // If this executor is busy, do not kill it unless we are told to
force kill it (SPARK-9552)
- val executorsToKill = knownExecutors
- .filter { id => !executorsPendingToRemove.contains(id) }
- .filter { id => force || !scheduler.isExecutorBusy(id) }
- executorsToKill.foreach { id => executorsPendingToRemove(id) =
!replace }
-
- // If we do not wish to replace the executors we kill, sync the target
number of executors
- // with the cluster manager to avoid allocating new ones. When
computing the new target,
- // take into account executors that are pending to be added or removed.
- if (!replace) {
- doRequestTotalExecutors(
- numExistingExecutors + numPendingExecutors -
executorsPendingToRemove.size)
- } else {
- numPendingExecutors += knownExecutors.size
+ val response = synchronized {
+ val (knownExecutors, unknownExecutors) =
executorIds.partition(executorDataMap.contains)
+ unknownExecutors.foreach { id =>
+ logWarning(s"Executor to kill $id does not exist!")
+ }
+
+ // If an executor is already pending to be removed, do not kill it
again (SPARK-9795)
+ // If this executor is busy, do not kill it unless we are told to
force kill it (SPARK-9552)
+ val executorsToKill = knownExecutors
+ .filter { id => !executorsPendingToRemove.contains(id) }
+ .filter { id => force || !scheduler.isExecutorBusy(id) }
+ executorsToKill.foreach { id => executorsPendingToRemove(id) =
!replace }
+
+ // If we do not wish to replace the executors we kill, sync the
target number of executors
+ // with the cluster manager to avoid allocating new ones. When
computing the new target,
+ // take into account executors that are pending to be added or
removed.
+ val adjustTotalExecutors =
+ if (!replace) {
+ doRequestTotalExecutors(
+ numExistingExecutors + numPendingExecutors -
executorsPendingToRemove.size)
+ } else {
+ numPendingExecutors += knownExecutors.size
+ Future.successful(true)
+ }
+
+ val killExecutors: Boolean => Future[Boolean] =
+ if (!executorsToKill.isEmpty) {
+ _ => doKillExecutors(executorsToKill)
+ } else {
+ _ => Future.successful(false)
+ }
+
+ adjustTotalExecutors.flatMap(killExecutors)(ThreadUtils.sameThread)
--- End diff --
I'm pretty sure you're correct, but at the same time I don't think there's
a requirement that `doKillExecutors` needs to be called from a synchronized
block. Current implementations just send RPC messages, which is probably better
done outside the synchronized block anyway.
---
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]
