This is an automated email from the ASF dual-hosted git repository.
gurwls223 pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/spark.git
The following commit(s) were added to refs/heads/master by this push:
new 17430fe4702 [SPARK-45302][PYTHON] Remove PID communication between
Python workers when no demon is used
17430fe4702 is described below
commit 17430fe47029f1d27c7913468b95abfd856fddcc
Author: Hyukjin Kwon <[email protected]>
AuthorDate: Wed Sep 27 10:48:17 2023 +0900
[SPARK-45302][PYTHON] Remove PID communication between Python workers when
no demon is used
### What changes were proposed in this pull request?
This PR removes the legacy workaround for JDK 8 in `PythonWorkerFactory`.
### Why are the changes needed?
No need to manually send the PID around through the socket.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
There are existing unittests for the daemon disabled.
### Was this patch authored or co-authored using generative AI tooling?
No.
Closes #43087 from HyukjinKwon/SPARK-45302.
Lead-authored-by: Hyukjin Kwon <[email protected]>
Co-authored-by: Hyukjin Kwon <[email protected]>
Signed-off-by: Hyukjin Kwon <[email protected]>
---
core/src/main/scala/org/apache/spark/SparkEnv.scala | 4 ++--
.../scala/org/apache/spark/api/python/PythonRunner.scala | 10 +++++-----
.../org/apache/spark/api/python/PythonWorkerFactory.scala | 15 +++++++--------
python/pyspark/daemon.py | 4 ++--
.../sql/connect/streaming/worker/foreach_batch_worker.py | 2 --
.../sql/connect/streaming/worker/listener_worker.py | 2 --
python/pyspark/sql/worker/analyze_udtf.py | 3 ---
python/pyspark/worker.py | 3 ---
.../spark/sql/execution/python/PythonArrowOutput.scala | 2 +-
.../spark/sql/execution/python/PythonUDFRunner.scala | 2 +-
10 files changed, 18 insertions(+), 29 deletions(-)
diff --git a/core/src/main/scala/org/apache/spark/SparkEnv.scala
b/core/src/main/scala/org/apache/spark/SparkEnv.scala
index e404c9ee8b4..937170b5ee8 100644
--- a/core/src/main/scala/org/apache/spark/SparkEnv.scala
+++ b/core/src/main/scala/org/apache/spark/SparkEnv.scala
@@ -128,7 +128,7 @@ class SparkEnv (
pythonExec: String,
workerModule: String,
daemonModule: String,
- envVars: Map[String, String]): (PythonWorker, Option[Int]) = {
+ envVars: Map[String, String]): (PythonWorker, Option[Long]) = {
synchronized {
val key = PythonWorkersKey(pythonExec, workerModule, daemonModule,
envVars)
pythonWorkers.getOrElseUpdate(key,
@@ -139,7 +139,7 @@ class SparkEnv (
private[spark] def createPythonWorker(
pythonExec: String,
workerModule: String,
- envVars: Map[String, String]): (PythonWorker, Option[Int]) = {
+ envVars: Map[String, String]): (PythonWorker, Option[Long]) = {
createPythonWorker(
pythonExec, workerModule, PythonWorkerFactory.defaultDaemonModule,
envVars)
}
diff --git a/core/src/main/scala/org/apache/spark/api/python/PythonRunner.scala
b/core/src/main/scala/org/apache/spark/api/python/PythonRunner.scala
index db95e6c2bd6..2a63298d0a1 100644
--- a/core/src/main/scala/org/apache/spark/api/python/PythonRunner.scala
+++ b/core/src/main/scala/org/apache/spark/api/python/PythonRunner.scala
@@ -84,7 +84,7 @@ private object BasePythonRunner {
private lazy val faultHandlerLogDir = Utils.createTempDir(namePrefix =
"faulthandler")
- private def faultHandlerLogPath(pid: Int): Path = {
+ private def faultHandlerLogPath(pid: Long): Path = {
new File(faultHandlerLogDir, pid.toString).toPath
}
}
@@ -200,7 +200,7 @@ private[spark] abstract class BasePythonRunner[IN, OUT](
envVars.put("SPARK_JOB_ARTIFACT_UUID",
jobArtifactUUID.getOrElse("default"))
- val (worker: PythonWorker, pid: Option[Int]) = env.createPythonWorker(
+ val (worker: PythonWorker, pid: Option[Long]) = env.createPythonWorker(
pythonExec, workerModule, daemonModule, envVars.asScala.toMap)
// Whether is the worker released into idle pool or closed. When any codes
try to release or
// close a worker, they should use `releasedOrClosed.compareAndSet` to
flip the state to make
@@ -253,7 +253,7 @@ private[spark] abstract class BasePythonRunner[IN, OUT](
startTime: Long,
env: SparkEnv,
worker: PythonWorker,
- pid: Option[Int],
+ pid: Option[Long],
releasedOrClosed: AtomicBoolean,
context: TaskContext): Iterator[OUT]
@@ -463,7 +463,7 @@ private[spark] abstract class BasePythonRunner[IN, OUT](
startTime: Long,
env: SparkEnv,
worker: PythonWorker,
- pid: Option[Int],
+ pid: Option[Long],
releasedOrClosed: AtomicBoolean,
context: TaskContext)
extends Iterator[OUT] {
@@ -838,7 +838,7 @@ private[spark] class PythonRunner(
startTime: Long,
env: SparkEnv,
worker: PythonWorker,
- pid: Option[Int],
+ pid: Option[Long],
releasedOrClosed: AtomicBoolean,
context: TaskContext): Iterator[Array[Byte]] = {
new ReaderIterator(
diff --git
a/core/src/main/scala/org/apache/spark/api/python/PythonWorkerFactory.scala
b/core/src/main/scala/org/apache/spark/api/python/PythonWorkerFactory.scala
index d0776eb2cc7..cef815b22ac 100644
--- a/core/src/main/scala/org/apache/spark/api/python/PythonWorkerFactory.scala
+++ b/core/src/main/scala/org/apache/spark/api/python/PythonWorkerFactory.scala
@@ -76,7 +76,7 @@ private[spark] class PythonWorkerFactory(
@GuardedBy("self")
private var daemonPort: Int = 0
@GuardedBy("self")
- private val daemonWorkers = new mutable.WeakHashMap[PythonWorker, Int]()
+ private val daemonWorkers = new mutable.WeakHashMap[PythonWorker, Long]()
@GuardedBy("self")
private val idleWorkers = new mutable.Queue[PythonWorker]()
@GuardedBy("self")
@@ -91,7 +91,7 @@ private[spark] class PythonWorkerFactory(
envVars.getOrElse("PYTHONPATH", ""),
sys.env.getOrElse("PYTHONPATH", ""))
- def create(): (PythonWorker, Option[Int]) = {
+ def create(): (PythonWorker, Option[Long]) = {
if (useDaemon) {
self.synchronized {
if (idleWorkers.nonEmpty) {
@@ -111,9 +111,9 @@ private[spark] class PythonWorkerFactory(
* processes itself to avoid the high cost of forking from Java. This
currently only works
* on UNIX-based systems.
*/
- private def createThroughDaemon(): (PythonWorker, Option[Int]) = {
+ private def createThroughDaemon(): (PythonWorker, Option[Long]) = {
- def createWorker(): (PythonWorker, Option[Int]) = {
+ def createWorker(): (PythonWorker, Option[Long]) = {
val socketChannel = SocketChannel.open(new InetSocketAddress(daemonHost,
daemonPort))
// These calls are blocking.
val pid = new
DataInputStream(Channels.newInputStream(socketChannel)).readInt()
@@ -153,7 +153,7 @@ private[spark] class PythonWorkerFactory(
/**
* Launch a worker by executing worker.py (by default) directly and telling
it to connect to us.
*/
- private[spark] def createSimpleWorker(blockingMode: Boolean): (PythonWorker,
Option[Int]) = {
+ private[spark] def createSimpleWorker(blockingMode: Boolean): (PythonWorker,
Option[Long]) = {
var serverSocketChannel: ServerSocketChannel = null
try {
serverSocketChannel = ServerSocketChannel.open()
@@ -189,8 +189,7 @@ private[spark] class PythonWorkerFactory(
try {
val socketChannel = serverSocketChannel.accept()
authHelper.authClient(socketChannel.socket())
- // TODO: When we drop JDK 8, we can just use workerProcess.pid()
- val pid = new
DataInputStream(Channels.newInputStream(socketChannel)).readInt()
+ val pid = workerProcess.toHandle.pid()
if (pid < 0) {
throw new IllegalStateException("Python failed to launch worker with
code " + pid)
}
@@ -386,7 +385,7 @@ private[spark] class PythonWorkerFactory(
daemonWorkers.get(worker).foreach { pid =>
// tell daemon to kill worker by pid
val output = new DataOutputStream(daemon.getOutputStream)
- output.writeInt(pid)
+ output.writeLong(pid)
output.flush()
daemon.getOutputStream.flush()
}
diff --git a/python/pyspark/daemon.py b/python/pyspark/daemon.py
index bbbc495d053..b0e06d13bed 100644
--- a/python/pyspark/daemon.py
+++ b/python/pyspark/daemon.py
@@ -28,7 +28,7 @@ from errno import EINTR, EAGAIN
from socket import AF_INET, AF_INET6, SOCK_STREAM, SOMAXCONN
from signal import SIGHUP, SIGTERM, SIGCHLD, SIG_DFL, SIG_IGN, SIGINT
-from pyspark.serializers import read_int, write_int, write_with_length,
UTF8Deserializer
+from pyspark.serializers import read_long, write_int, write_with_length,
UTF8Deserializer
if len(sys.argv) > 1:
import importlib
@@ -139,7 +139,7 @@ def manager():
if 0 in ready_fds:
try:
- worker_pid = read_int(stdin_bin)
+ worker_pid = read_long(stdin_bin)
except EOFError:
# Spark told us to exit by closing stdin
shutdown(0)
diff --git
a/python/pyspark/sql/connect/streaming/worker/foreach_batch_worker.py
b/python/pyspark/sql/connect/streaming/worker/foreach_batch_worker.py
index 06534e355de..022e768c43b 100644
--- a/python/pyspark/sql/connect/streaming/worker/foreach_batch_worker.py
+++ b/python/pyspark/sql/connect/streaming/worker/foreach_batch_worker.py
@@ -87,6 +87,4 @@ if __name__ == "__main__":
(sock_file, sock) = local_connect_and_auth(java_port, auth_secret)
# There could be a long time between each micro batch.
sock.settimeout(None)
- write_int(os.getpid(), sock_file)
- sock_file.flush()
main(sock_file, sock_file)
diff --git a/python/pyspark/sql/connect/streaming/worker/listener_worker.py
b/python/pyspark/sql/connect/streaming/worker/listener_worker.py
index ed38a788435..bb6bcd5d965 100644
--- a/python/pyspark/sql/connect/streaming/worker/listener_worker.py
+++ b/python/pyspark/sql/connect/streaming/worker/listener_worker.py
@@ -101,6 +101,4 @@ if __name__ == "__main__":
(sock_file, sock) = local_connect_and_auth(java_port, auth_secret)
# There could be a long time between each listener event.
sock.settimeout(None)
- write_int(os.getpid(), sock_file)
- sock_file.flush()
main(sock_file, sock_file)
diff --git a/python/pyspark/sql/worker/analyze_udtf.py
b/python/pyspark/sql/worker/analyze_udtf.py
index 194cd3db765..6fb3ca995e5 100644
--- a/python/pyspark/sql/worker/analyze_udtf.py
+++ b/python/pyspark/sql/worker/analyze_udtf.py
@@ -164,7 +164,4 @@ if __name__ == "__main__":
java_port = int(os.environ["PYTHON_WORKER_FACTORY_PORT"])
auth_secret = os.environ["PYTHON_WORKER_FACTORY_SECRET"]
(sock_file, _) = local_connect_and_auth(java_port, auth_secret)
- # TODO: Remove the following two lines and use `Process.pid()` when we
drop JDK 8.
- write_int(os.getpid(), sock_file)
- sock_file.flush()
main(sock_file, sock_file)
diff --git a/python/pyspark/worker.py b/python/pyspark/worker.py
index a3c7bbb59dd..77481704979 100644
--- a/python/pyspark/worker.py
+++ b/python/pyspark/worker.py
@@ -1352,7 +1352,4 @@ if __name__ == "__main__":
java_port = int(os.environ["PYTHON_WORKER_FACTORY_PORT"])
auth_secret = os.environ["PYTHON_WORKER_FACTORY_SECRET"]
(sock_file, _) = local_connect_and_auth(java_port, auth_secret)
- # TODO: Remove the following two lines and use `Process.pid()` when we
drop JDK 8.
- write_int(os.getpid(), sock_file)
- sock_file.flush()
main(sock_file, sock_file)
diff --git
a/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonArrowOutput.scala
b/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonArrowOutput.scala
index 8f99325e4e0..2e410eae61e 100644
---
a/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonArrowOutput.scala
+++
b/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonArrowOutput.scala
@@ -49,7 +49,7 @@ private[python] trait PythonArrowOutput[OUT <: AnyRef] {
self: BasePythonRunner[
startTime: Long,
env: SparkEnv,
worker: PythonWorker,
- pid: Option[Int],
+ pid: Option[Long],
releasedOrClosed: AtomicBoolean,
context: TaskContext): Iterator[OUT] = {
diff --git
a/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonUDFRunner.scala
b/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonUDFRunner.scala
index b99517f544d..12c51506b13 100644
---
a/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonUDFRunner.scala
+++
b/sql/core/src/main/scala/org/apache/spark/sql/execution/python/PythonUDFRunner.scala
@@ -78,7 +78,7 @@ abstract class BasePythonUDFRunner(
startTime: Long,
env: SparkEnv,
worker: PythonWorker,
- pid: Option[Int],
+ pid: Option[Long],
releasedOrClosed: AtomicBoolean,
context: TaskContext): Iterator[Array[Byte]] = {
new ReaderIterator(
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
