Skip to content

[SPARK-37688][CORE] ExecutorMonitor should ignore SparkListenerBlockUpdated event if executor was not active #34956

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
sleep1661 wants to merge 2 commits into from
Closed

[SPARK-37688][CORE] ExecutorMonitor should ignore SparkListenerBlockUpdated event if executor was not active #34956

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
94612bf
[SPARK-37688]ExecutorMonitor should ignore SparkListenerBlockUpdated …
sleep1661 Nov 12, 2021
289717a
fix up unit test
sleep1661 Dec 20, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions core/src/main/scala/org/apache/spark/scheduler/dynalloc/ExecutorMonitor.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -378,6 +378,10 @@ private[spark] class ExecutorMonitor(
}

override def onBlockUpdated(event: SparkListenerBlockUpdated): Unit = {
if (!client.isExecutorActive(event.blockUpdatedInfo.blockManagerId.executorId)) {
return
}

val exec = ensureExecutorIsTracked(event.blockUpdatedInfo.blockManagerId.executorId,
UNKNOWN_RESOURCE_PROFILE_ID)

Expand Down
19 changes: 19 additions & 0 deletions core/src/test/scala/org/apache/spark/scheduler/dynalloc/ExecutorMonitorSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -169,6 +169,8 @@ class ExecutorMonitorSuite extends SparkFunSuite {
}

test("keeps track of stored blocks for each rdd and split") {
knownExecs ++= Set("1", "2")

monitor.onExecutorAdded(SparkListenerExecutorAdded(clock.getTimeMillis(), "1", execInfo))

monitor.onBlockUpdated(rddUpdate(1, 0, "1"))
Expand Down Expand Up @@ -234,6 +236,7 @@ class ExecutorMonitorSuite extends SparkFunSuite {
}

test("SPARK-27677: don't track blocks stored on disk when using shuffle service") {
knownExecs += "1"
// First make sure that blocks on disk are counted when no shuffle service is available.
monitor.onExecutorAdded(SparkListenerExecutorAdded(clock.getTimeMillis(), "1", execInfo))
monitor.onBlockUpdated(rddUpdate(1, 0, "1", level = StorageLevel.DISK_ONLY))
Expand Down Expand Up @@ -443,6 +446,22 @@ class ExecutorMonitorSuite extends SparkFunSuite {
assert(monitor.timedOutExecutors(idleDeadline).isEmpty)
}

test("SPARK-37688: ignore SparkListenerBlockUpdated event if executor was not active") {
conf
.set(DYN_ALLOCATION_SHUFFLE_TRACKING_TIMEOUT, Long.MaxValue)
.set(DYN_ALLOCATION_SHUFFLE_TRACKING_ENABLED, true)
.set(SHUFFLE_SERVICE_ENABLED, false)
monitor = new ExecutorMonitor(conf, client, null, clock, allocationManagerSource())

monitor.onExecutorAdded(SparkListenerExecutorAdded(clock.getTimeMillis(), "1", execInfo))
monitor.onExecutorRemoved(SparkListenerExecutorRemoved(clock.getTimeMillis(), "1",
"heartbeats timeout"))
monitor.onBlockUpdated(rddUpdate(1, 1, "1", level = StorageLevel.MEMORY_AND_DISK))

assert(monitor.executorCount == 0 )
}


private def idleDeadline: Long = clock.nanoTime() + idleTimeoutNs + 1
private def storageDeadline: Long = clock.nanoTime() + storageTimeoutNs + 1
private def shuffleDeadline: Long = clock.nanoTime() + shuffleTimeoutNs + 1
Expand Down