Skip to content

[SPARK-22837][SQL]Session timeout checker does not work in SessionManager. #20025

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
zuotingbing wants to merge 4 commits into from
Closed

[SPARK-22837][SQL]Session timeout checker does not work in SessionManager. #20025

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
34b7680
[SPARK-22837][SQL]Session timeout checker does not work in SessionMan…
Dec 20, 2017
79fc11a
fix code review
Jan 17, 2018
5b3c06f
Keep createBackgroundOperationPool unchanged. Looks like the hive thr…
Jan 19, 2018
883deb2
fix code review
Jan 23, 2018
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
36 changes: 8 additions & 28 deletions sql/hive-thriftserver/src/main/java/org/apache/hive/service/cli/session/SessionManager.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,11 +23,7 @@
import java.util.ArrayList;
import java.util.Date;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.Future;
import java.util.concurrent.LinkedBlockingQueue;
import java.util.concurrent.ThreadPoolExecutor;
import java.util.concurrent.TimeUnit;
Copy link
Member

@gatorsmile gatorsmile Jan 19, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

revert this back.

Copy link
Author

@zuotingbing zuotingbing Jan 23, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes, thanks

import java.util.concurrent.*;

import org.apache.commons.io.FileUtils;
import org.apache.commons.logging.Log;
Expand Down Expand Up @@ -79,35 +75,19 @@ public synchronized void init(HiveConf hiveConf) {
initOperationLogRootDir();
}
createBackgroundOperationPool();
addService(operationManager);
super.init(hiveConf);
Copy link

@liufengdb liufengdb Jan 19, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

hmm, I think we revert keep this line too.

Copy link
Author

@zuotingbing zuotingbing Jan 23, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes, we could keep super.init() here and remove initCompositeService in SparkSQLSessionManager.init(). thanks

}

private void createBackgroundOperationPool() {
int poolSize = hiveConf.getIntVar(ConfVars.HIVE_SERVER2_ASYNC_EXEC_THREADS);
LOG.info("HiveServer2: Background operation thread pool size: " + poolSize);
int poolQueueSize = hiveConf.getIntVar(ConfVars.HIVE_SERVER2_ASYNC_EXEC_WAIT_QUEUE_SIZE);
LOG.info("HiveServer2: Background operation thread wait queue size: " + poolQueueSize);
long keepAliveTime = HiveConf.getTimeVar(
hiveConf, ConfVars.HIVE_SERVER2_ASYNC_EXEC_KEEPALIVE_TIME, TimeUnit.SECONDS);
LOG.info(
"HiveServer2: Background operation thread keepalive time: " + keepAliveTime + " seconds");

// Create a thread pool with #poolSize threads
// Threads terminate when they are idle for more than the keepAliveTime
// A bounded blocking queue is used to queue incoming operations, if #operations > poolSize
String threadPoolName = "HiveServer2-Background-Pool";
backgroundOperationPool = new ThreadPoolExecutor(poolSize, poolSize,
keepAliveTime, TimeUnit.SECONDS, new LinkedBlockingQueue<Runnable>(poolQueueSize),
new ThreadFactoryWithGarbageCleanup(threadPoolName));
backgroundOperationPool.allowCoreThreadTimeOut(true);

Copy link

@liufengdb liufengdb Jan 18, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think we can keep this file unchanged. Looks like the hive thread pool is more flexible than the spark one.

Copy link
Author

@zuotingbing zuotingbing Jan 19, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes, i will update it . Thanks!

checkInterval = HiveConf.getTimeVar(
hiveConf, ConfVars.HIVE_SERVER2_SESSION_CHECK_INTERVAL, TimeUnit.MILLISECONDS);
sessionTimeout = HiveConf.getTimeVar(
hiveConf, ConfVars.HIVE_SERVER2_IDLE_SESSION_TIMEOUT, TimeUnit.MILLISECONDS);
checkOperation = HiveConf.getBoolVar(hiveConf,
ConfVars.HIVE_SERVER2_IDLE_SESSION_CHECK_OPERATION);
addService(operationManager);
}

private void createBackgroundOperationPool() {
int backgroundPoolSize = hiveConf.getIntVar(ConfVars.HIVE_SERVER2_ASYNC_EXEC_THREADS);
backgroundOperationPool = (ThreadPoolExecutor) Executors.newFixedThreadPool(backgroundPoolSize);
LOG.info("HiveServer2: Async execution pool size: " + backgroundPoolSize);
}

private void initOperationLogRootDir() {
Expand Down
15 changes: 1 addition & 14 deletions ...server/src/main/scala/org/apache/spark/sql/hive/thriftserver/SparkSQLSessionManager.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,21 +40,8 @@ private[hive] class SparkSQLSessionManager(hiveServer: HiveServer2, sqlContext:
private lazy val sparkSqlOperationManager = new SparkSQLOperationManager()

override def init(hiveConf: HiveConf) {
setSuperField(this, "hiveConf", hiveConf)

// Create operation log root directory, if operation logging is enabled
if (hiveConf.getBoolVar(ConfVars.HIVE_SERVER2_LOGGING_OPERATION_ENABLED)) {
invoke(classOf[SessionManager], this, "initOperationLogRootDir")
}

val backgroundPoolSize = hiveConf.getIntVar(ConfVars.HIVE_SERVER2_ASYNC_EXEC_THREADS)
setSuperField(this, "backgroundOperationPool", Executors.newFixedThreadPool(backgroundPoolSize))
getAncestorField[Log](this, 3, "LOG").info(
s"HiveServer2: Async execution pool size $backgroundPoolSize")

setSuperField(this, "operationManager", sparkSqlOperationManager)
addService(sparkSqlOperationManager)

super.init(hiveConf)
initCompositeService(hiveConf)
}

Expand Down