Thread Producer-Consumer java - java

i have to finish an exercise where i have to go find ".java" files in my folder path using the producer/consumer pattern with at least one producer thread and x consumer threads.
ProducerConsumer-class:
First i tried to stop the consumer when the producer is finished finding files with setting a while loop from true to false which doesn't work. It doesn't work because the threads are still running obviously just not doing anything useful. Now i use a closePool() function (as well).
So the function does work if i dont put up with my locks called locka. And thats basically something i don't understand.
So if i have
loka.lock();
ende = false;
loka.unlock();
and
while(ende){
loka.lock();
System.out.println(xy.getQueue());
loka.unlock();
}
the closePool() function will never get called. And this is something i don't understand. If i put away the locks in the while loop it does work and the threads do stop.
questions:
1) The ende parameter will be set false anyway so the lock will be finally released.
2) Secondly i did only lock a part of the method and not the object?! As far as i understand it other code in other methods in the same object will still work at the same time. Or is the lock like synchronized and i synchronize the whole object while it is in the lock state?
In my understanding the while loop in the consumer-thread is locked but the producer-thread will still call closePool();
on a extra note: maybe i didn't even design my Producer/Consumer pattern the right way.
import java.io.IOException;
import java.nio.file.FileVisitResult;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.nio.file.SimpleFileVisitor;
import java.nio.file.attribute.BasicFileAttributes;
import java.util.concurrent.ArrayBlockingQueue;
import java.util.concurrent.BlockingQueue;
public class FindJavaVisitorp extends SimpleFileVisitor<Path> {
private BlockingQueue<String> xxx = new ArrayBlockingQueue<String>(10);
public FileVisitResult visitFile(Path file, BasicFileAttributes attrs) {
if (file.toString().endsWith(".java")) {
try {
xxx.put(file.toString());
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
return FileVisitResult.CONTINUE;
}
public String getQueue() throws InterruptedException {
return xxx.take();
}
}
public class ProducerConsumer {
private volatile boolean ende = true;
private Path path;
private FindJavaVisitorp xy;
private Lock loka = new ReentrantLock();
private ExecutorService pepe;
public ProducerConsumer(Path path, FindJavaVisitorp xy, ExecutorService xyz) {
this.path = path;
this.xy = xy;
pepe = xyz;
}
public void produce() throws IOException, InterruptedException {
Files.walkFileTree(path, xy);
loka.lock();
ende = false;
loka.unlock();
closePool();
}
public void consume() throws InterruptedException {
while (ende) {
loka.lock();
System.out.println(xy.getQueue());
loka.unlock();
}
}
public void closePool() {
pepe.shutdown();
try {
if (!pepe.awaitTermination(60, TimeUnit.SECONDS)) {
pepe.shutdownNow();
if (!pepe.awaitTermination(60, TimeUnit.SECONDS)) {
System.err.println("Pool couldn't be terminated!");
}
}
} catch (InterruptedException e) {
pepe.shutdownNow();
}
}
}
public class Test {
public static void main(String[] args) {
Path startingDir = Paths.get("/usr/local/");
FindJavaVisitorp x = new FindJavaVisitorp();
ExecutorService exec = Executors.newCachedThreadPool();
final ProducerConsumer pp = new ProducerConsumer(startingDir, x, exec);
exec.submit(new Runnable() {
public void run() {
try {
pp.produce();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
});
// x.printQueue();
for (int j = 0; j < 5; j++) {
exec.submit(new Runnable() {
public void run() {
try {
pp.consume();
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
});
}
exec.shutdown();
}
}

Yes, but only if thread goes to check it and not waiting for something else, which in your case is happening. In the while loop threads are waiting for BlockingQueue and if queue is empty then then will not get a chance to check the flag variable. Also you don't need locks as you're already using BlockingQueue. In your example there's no relation between the two critical sections.
.
Following code will try to guard loka=false from any concurrent access.
loka.lock();
ende = false;//critical section
loka.unlock();
Following code will be free from concurrent access and mutually exclusive from the above critical section.
while(ende){
loka.lock();
System.out.println(xy.getQueue());//critical section
loka.unlock();
}
As there's nothing common between these two critical sections, mutual exclusion is doing nothing. Since ende is volatile guarding it with locks doesn't do anything as primitive types already have atomic access.
Reads and writes are atomic for reference variables and for most primitive variables (all types except long and double).
Reads and writes are atomic for all variables declared volatile (including long and double variables).
Only code inside the guarded block by lock() andunlock()` will be locked from concurrent access. Object itself is free to do any concurrent simultaneous (to the locked block) task outside these blocks.
And finally follow proper naming conventions and give your variables meaningful names.
Main answer to your problem why your threads are still running is because they're waiting on the blockingQueue.takeItem() and they can not be released from it unless queue is filled again, however since Producer is finished there's no possibility of that happening.
How to avoid this behavior
There are no methods on BlockingQueue which allow immediate release of waiting threads
One thing we can do is make producer put a LAST_ITEM and have consumers check if the item they got is LAST_ITEM and thus they can release themselves.
Following is working code. I have made some modifications to the variable and method names to make them more meaningful.
JavaFileVisitor
package filevisitor;
import java.nio.file.FileVisitResult;
import java.nio.file.Path;
import java.nio.file.SimpleFileVisitor;
import java.nio.file.attribute.BasicFileAttributes;
import java.util.concurrent.ArrayBlockingQueue;
import java.util.concurrent.BlockingQueue;
public class JavaFileVisitor extends SimpleFileVisitor<Path> {
private BlockingQueue<String> blockingQueue = new ArrayBlockingQueue<String>(10);
public static String NO_MORE_ITEMS = "### NO MORE ITEMS ###";
public FileVisitResult visitFile(Path file, BasicFileAttributes attrs) {
if (file.toString().endsWith(".java")) {
try {
blockingQueue.put(file.toString());
} catch (InterruptedException e) {
e.printStackTrace();
}
}
return FileVisitResult.CONTINUE;
}
public String getQueueItem() throws InterruptedException {
String item = blockingQueue.take();
if(NO_MORE_ITEMS.equals(item)) {
setNoMoreItems();
}
return item;
}
public void setNoMoreItems() {
try {
blockingQueue.put(NO_MORE_ITEMS);
} catch (InterruptedException e) {
e.printStackTrace();
}
}
}
ProducerConsumer
package filevisitor;
import java.io.IOException;
import java.nio.file.Files;
import java.nio.file.Path;
public class ProducerConsumer {
private Path path;
private JavaFileVisitor fileVisitor;
public ProducerConsumer(Path path, JavaFileVisitor visitor) {
this.path = path;
this.fileVisitor = visitor;
}
public void produce() throws IOException, InterruptedException {
Files.walkFileTree(path, fileVisitor);
fileVisitor.setNoMoreItems();
}
public void consume() throws InterruptedException {
while (true) {
String item = fileVisitor.getQueueItem();
if(JavaFileVisitor.NO_MORE_ITEMS.equals(item)) {
break;
}
System.out.println(item);
}
}
}
ProducerConsumerMain
package filevisitor;
import java.io.IOException;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.TimeUnit;
public class ProducerConsumerMain {
public static void main(String[] args) {
Path startingDir = Paths.get("src/filevisitor");
JavaFileVisitor fileVisitor = new JavaFileVisitor();
ExecutorService executor = Executors.newCachedThreadPool();
final ProducerConsumer producerConsumer = new ProducerConsumer(startingDir, fileVisitor);
executor.submit(new Runnable() {
public void run() {
System.out.println("Producer started");
try {
producerConsumer.produce();
} catch (IOException e) {
e.printStackTrace();
} catch (InterruptedException e) {
e.printStackTrace();
}
System.out.println("Producer finished");
}
});
for (int j = 0; j < 5; j++) {
executor.submit(new Runnable() {
public void run() {
String threadName = Thread.currentThread().getName();
System.out.println(threadName + " Consumer Started");
try {
producerConsumer.consume();
} catch (InterruptedException e) {
e.printStackTrace();
}
System.out.println(threadName + " Consumer finished");
}
});
}
executor.shutdown();
System.out.println("Executor shutdown, waiting for threads to finish");
try {
executor.awaitTermination(60, TimeUnit.SECONDS);
} catch (InterruptedException e) {
e.printStackTrace();
}
System.out.println("Exiting main");
}
}
Output
Producer started
pool-1-thread-3 Consumer Started
pool-1-thread-2 Consumer Started
Executor shutdown, waiting for threads to finish
pool-1-thread-5 Consumer Started
pool-1-thread-6 Consumer Started
pool-1-thread-4 Consumer Started
src\filevisitor\JavaFileVisitor.java
src\filevisitor\ProducerConsumerMain.java
src\filevisitor\ProducerConsumer.java
pool-1-thread-6 Consumer finished
pool-1-thread-4 Consumer finished
pool-1-thread-3 Consumer finished
pool-1-thread-5 Consumer finished
Producer finished
pool-1-thread-2 Consumer finished
Exiting main

Related

How does Java non-fair ReentrantReadWriteLock really work?

I have written the following test code in Java using ReentrantReadWriteLock to understand the difference between fair and non-fair mode. However, I see in both modes the result and output is always the same. It seems it's always working in fair mode. Can anybody explain in which case fair and non-fair mode will result in different behaviors?
package lockTest;
import java.util.concurrent.locks.ReadWriteLock;
import java.util.concurrent.locks.ReentrantReadWriteLock;
public class MyLockTest {
static private ReadWriteLock myLock = new ReentrantReadWriteLock(false);
public class Reader extends Thread {
int val_;
public Reader(int val) {
val_ = val;
}
public void run() {
if (val_ > 0) {
try {
Thread.sleep(3000);
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
myLock.readLock().lock();
System.out.println(Thread.currentThread().getName() + ": Reader inside critical section - val: " + val_ + "-----");
try {
Thread.sleep(6000);
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
myLock.readLock().unlock();
}
}
public class Writer extends Thread {
public void run() {
try {
Thread.sleep(1000);
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
myLock.writeLock().lock();
System.out.println(Thread.currentThread().getName() + ": Writer inside critical section *****");
try {
Thread.sleep(6000);
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
myLock.writeLock().unlock();
}
}
public static void main(String[] args) {
MyLockTest test1 = new MyLockTest();
MyLockTest.Reader reader1 = test1.new Reader(0);
MyLockTest.Writer writer1 = test1.new Writer();
MyLockTest.Reader reader2 = test1.new Reader(1);
reader2.start();
writer1.start();
reader1.start();
}
}
The output is always:
Thread-0: Reader inside critical section - val: 0-----
Thread-1: Writer inside critical section *****
Thread-2: Reader inside critical section - val: 1-----
The output above is what I expect to see when I change Lock creation to fair mode:
static private ReadWriteLock myLock = new ReentrantReadWriteLock(true);
For non-fair mode I would expect to see the following output:
Thread-0: Reader inside critical section - val: 0-----
Thread-2: Reader inside critical section - val: 1-----
Thread-1: Writer inside critical section *****
Using "fair" versus "non-fair" mode affects how the lock is assigned to threads in case of contention.
From the Javadoc for ReentrantReadWriteLock: using the "non-fair" mode, the order of entry to the read and write lock is unspecified, while using the "fair" mode, threads contend for entry using an approximately arrival-order policy.
We can see how using fair/non-fair affects program execution by having some thread contend over the same lock; see the program below.
Running the sample code, a ReentrantWriteLock is contended by different threads; after 1000 lock operations, we dump how many times each thread acquired the lock.
In case USE_FAIR=false is used, counts are random, and a possible output is:
Thread thread-B finished, count=920
Thread thread-A finished, count=79
Thread thread-D finished, count=0
Thread thread-C finished, count=0
in case USE_FAIR=true is used, the output is always like
Thread thread-D finished, count=249
Thread thread-A finished, count=250
Thread thread-C finished, count=250
Thread thread-B finished, count=250
Sample code
package sample1;
import java.util.concurrent.locks.ReentrantReadWriteLock;
public class UseLock {
public static void main(String[] args) {
UseLock o = new UseLock();
o.go();
}
private void go() {
TotalPermits tp = new TotalPermits();
tp.lock.writeLock().lock();
Contender a = new Contender(tp, "thread-A");
Contender b = new Contender(tp, "thread-B");
Contender c = new Contender(tp, "thread-C");
Contender d = new Contender(tp, "thread-D");
a.start();
b.start();
c.start();
d.start();
tp.lock.writeLock().unlock();
}
}
class TotalPermits {
private static final boolean USE_FAIR = true;
private int count = 1_000;
ReentrantReadWriteLock lock = new ReentrantReadWriteLock(USE_FAIR);
public boolean get() {
try {
lock.writeLock().lock();
try {
Thread.sleep(1);
} catch (InterruptedException e) { }
return --count>0;
} finally {
lock.writeLock().unlock();
}
}
}
class Contender extends Thread {
private int count = 0;
final String name;
final TotalPermits tp;
Contender(TotalPermits tp, String name) {
this.tp = tp;
this.name = name;
}
#Override
public void run() {
while ( tp.get() ) {
count++;
}
System.out.printf("Thread %s finished, count=%d%n", name, count);
}
}
Note:
The sample code above uses a "write" lock, which can only be held by one thread at a time. So we can use that to divide N permits across the contenders. On the other hand, the "read" lock can be held by multiple threads, as long as none is holding the write lock.

Ski lift with capacity equals N.

I am learning about multithreading. It's my first task. I wrote this code and i can't move on. Task:
Ski lift with capacity equal N.
Clients have a weight (random Ki value) and They are threads that execute in
loop:
downhill(sleep(big random value)
Try to get into the lift (if the total weight of customers is
Less than or equal to N).
If it failed - they are waiting (sleep(small random value)
and re-execute the previous point.
if it was successful - they go up.
public class Client extends Thread
{
private SkiLift lift;
private int weight;
public Client(SkiLift l, int w)
{
this.lift = l;
this.weight=w;
}
public int getWeight()
{
return weight;
}
public void run()
{
for (int i =0; i<10; i++)
{
lift.downhill(this);
lift.goIn(this);
this.setPriority(MAX_PRIORITY);
lift.drive(this);
lift.goOut(this);
this.setPriority(5);
}
}
}
public class SkiLift
{
private static int actualLoad=0;
private static final int CAPACITY=300;
synchronized public void goIn(Client client)
{
try
{
System.out.println("Client " + client.getId() + " try to get into the lift");
while (actualLoad>CAPACITY)
{
System.out.println("The Lift is full!");
client.sleep((long) (Math.random()*1000));
wait();
}
}
catch (InterruptedException e) {}
System.out.println("Client " + client.getId() + "get into the lift " );
actualLoad+=client.getWeight();
System.out.println("actual load = " + actualLoad);
}
synchronized public void goOut (Client client)
{
System.out.println("Client "+ client.getId() + " leave the lift ");
actualLoad-=client.getWeight();
System.out.println("Actual load = " + actualLoad);
notifyAll();
}
public void downhill(Client client)
{
System.out.println("Client nr: " + client.getId()+ " downhill ");
try
{
client.sleep((long) (Math.random()*10000));
}
catch (InterruptedException e){}
}
public void drive(Client client)
{
try
{
client.sleep(9000);
}
catch (InterruptedException e){e.printStackTrace();}
}
}
I have three problems and i can't solve them:
The first who will enter must to be the first who has attempted to enter. (Just like in a queue)
The client who first came on the lift must also be the first to go down.
What is the moniotor in my program?
Thanks in advance :)
I think this question belongs to Codereview
Your Client should have a state like "topOfTheMountainReached", "liftStationReached", "liftEntered", ...
Your Client then waits for this events to happen. That's also the answer to your question which element to monitor - the state, or the client itself.
For the queue you can use a ArrayListBlockingQueue.
Your SkiLift then has to wait for new Clients to arrive and put them into the lift. As soon the client enters the lift, the client also gets notified that it has entered the lift. The Lift also notifies the client when the top is reached.
Here is an example of how such solution could look like.
It uses the Java Executor Service to schedule the events for getting the client out of the lift and for reaching the lift station at the end oft the downhill part. This may also be solved differently.
The Client:
import java.util.Random;
import java.util.concurrent.Executors;
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.TimeUnit;
public class Client implements Runnable{
final ScheduledExecutorService dhexceutors = Executors.newScheduledThreadPool(500);
final static Random DHRANDOM = new Random();
final long weight;
public enum State {
goDownhill,
waitForLift,
goUp,
onTop,
}
private State state;
public SkiLift lift;
public Client(long weight,SkiLift lift) {
this.lift = lift;
this.weight = weight;
this.state = State.onTop;
goDownHill();
}
private void enterLift() {
lift.add(this);
}
private void goDownHill() {
synchronized (this) {
state = State.goDownhill;
this.notify();
}
dhexceutors.schedule(() -> {
liftStationReached();
}, DHRANDOM.nextInt(500), TimeUnit.MILLISECONDS);
}
public void liftStationReached() {
synchronized(this) {
state = State.waitForLift;
this.notify();
}
}
public void topReached() {
synchronized(this) {
state = State.onTop;
this.notify();
}
}
public void liftEntered() {
synchronized(this) {
state = State.goUp;
this.notify();
}
}
public void run() {
synchronized(this) {
while (true) {
try {
this.wait();
switch (state) {
case waitForLift:
enterLift();
break;
case goUp:
// just wait for the topReached event
break;
case goDownhill:
// just wait for reaching the lift.
break;
case onTop:
goDownHill();
break;
}
} catch (InterruptedException ie) {
Thread.currentThread().interrupt();
}
}
}
}
}
The Lift:
package skilift;
import java.util.ArrayList;
import java.util.concurrent.ArrayBlockingQueue;
import java.util.concurrent.Executors;
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.TimeUnit;
public class SkiLift implements Runnable{
private ScheduledExecutorService getOutClientExecutor;
public SkiLift() {
getOutClientExecutor = Executors.newScheduledThreadPool(50);
waitingClientsQueue = new ArrayBlockingQueue<>(1000);
occupiedSeats = new ArrayList<>();
}
private final ArrayList<Client> occupiedSeats;
private long usedCapacity;
private final ArrayBlockingQueue<Client> waitingClientsQueue;
private final long capacity = 500;
public void add(Client client) {
synchronized(waitingClientsQueue) {
waitingClientsQueue.add(client);
waitingClientsQueue.notify();
}
}
private synchronized void occupySeat(Client client) {
occupiedSeats.add(client);
usedCapacity += client.weight;
}
private synchronized void getClientOut(Client client) {
occupiedSeats.remove(client);
usedCapacity -= client.weight;
// notify the waitingClientQueue that the capacity has changed
synchronized (waitingClientsQueue) {
waitingClientsQueue.notify();
}
client.topReached();
}
public void run() {
while (true) {
synchronized(waitingClientsQueue) {
try {
if (!waitingClientsQueue.isEmpty()) {
Client c = waitingClientsQueue.peek();
if (usedCapacity + c.weight <= capacity) {
occupySeat(waitingClientsQueue.poll());
getOutClientExecutor.schedule(() -> {
getClientOut(c);
}, 2, TimeUnit.SECONDS);
} else {
waitingClientsQueue.wait();
}
} else {
waitingClientsQueue.wait();
}
} catch (InterruptedException ie) {
Thread.currentThread().interrupt();
}
}
}
}
}
Apparently, the bottleneck in your system is the lift. You can only have N concurrent users of the lift.
Also, 3. mentions a Monitor. After some reading what a monitor is, you should figure out that it allows exclusive access to the limited resource, the lift.
So design your lift access to try to acquire one of the N monitors, wait a while, and at the end do not forget to release the monitor, so someone else can get it.

Future.isDone() does not show state change when using executor.scheduleAtFixedRate method

Hi I'm a newbie to concurrency so I wrote a very basic program to see whether on a threads completion the future.isDone() method shows true, unfortunately it always shows "false" when I schedule the task with scheduledAtFixedRate method. However if I use schedule method it shows "true" of course the simple task does not rerun seconds later. Any suggestions or explanations to help me understand why this is the case would be much appreciated.
Thanks!
Here with the code:
package com.company;
import org.testng.annotations.Test;
import java.util.ArrayList;
import java.util.List;
import java.util.concurrent.*;
public class Main {
public static void main(String[] args) {
Main mn = new Main();
System.out.println("Main thread started...");
mn.runobjects();
System.out.println("Main thread stopping...");
}
#Test
public void runobjects(){
List<commonrun> Obj = new ArrayList<>();
Obj.add(new TestObj1());
Obj.add(new TestObj2());
Obj.add(new TestObj3());
ScheduledExecutorService executor = Executors.newScheduledThreadPool(5);
ScheduledFuture<?> futures1 = null;
ScheduledFuture<?> futures2 = null;
ScheduledFuture<?> futures3 = null;
int i=0;
for (commonrun obj : Obj){
if (i==0) {
futures1 = executor.schedule(() -> obj.runme(), 0, TimeUnit.SECONDS);
}
if (i==1) {
futures2 = executor.scheduleAtFixedRate(() -> obj.runme(), 0, 10, TimeUnit.SECONDS);
}
if (i==2) {
futures3 = executor.scheduleAtFixedRate(() -> obj.runme(), 0, 10, TimeUnit.SECONDS);
}
i++;
}
while(true){
try {
Thread.sleep(1000);
} catch (InterruptedException e) {
e.printStackTrace();
}
System.out.println("Thread 1 is done : "+ futures1.isDone());
System.out.println("Thread 2 is done : "+ futures2.isDone());
System.out.println("Thread 3 is done : "+ futures3.isDone());
try {
Thread.sleep(1000);
} catch (InterruptedException e) {
e.printStackTrace();
}
}
}
}
package com.company;
public interface commonrun {
public void runme();
}
package com.company;
public class TestObj1 implements commonrun {
static int counter = 0;
#Override
public void runme() {
System.out.println("Object 1 Starting... run : " + counter);
try {
Thread.sleep(1000);
} catch (InterruptedException e) {
e.printStackTrace();
}
System.out.println("Object 1 Stopping... run : " + counter);
counter++;
}
}
package com.company;
public class TestObj2 implements commonrun {
#Override
public void runme() {
System.out.println("Object 2 Starting...");
try {
Thread.sleep(2000);
} catch (InterruptedException e) {
e.printStackTrace();
}
System.out.println("Object 2 Stopping...");
}
}
package com.company;
public class TestObj3 implements commonrun {
#Override
public void runme() {
System.out.println("Object 3 Starting...");
try {
Thread.sleep(3000);
} catch (InterruptedException e) {
e.printStackTrace();
}
System.out.println("Object 3 Stopping...");
}
}
ScheduledExecutorService.scheduleAtFixedRate() schedules repetitive task. It will never (naturally) finish, so it won't become done. From method's documentation:
the task will only terminate via cancellation or termination of the
executor
Thus, only if you call future.cancel() or executor.terminate() will the task become done and future.isDone() will then return true.
While one could potentially expect that future becomes done as soon as first task's execution completes, this is not the case for the following reasons:
once future becomes done it cannot be "undone" ("done" is a terminal state for a future), so isDone cannot report current execution state of a repetitive job
once future becomes done it makes no sense to cancel it (there isn't anything to cancel) -- that would not fit a repetitive task, which won't run indefinitely until canceled.

Why my threads are not completing

I am trying to understand countDownLatch and I have this program but I do not know why my program is not returning and not finishing.
package countDownLatches;
import java.util.concurrent.CountDownLatch;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
class Processor implements Runnable {
CountDownLatch latch;
public Processor(CountDownLatch latch) {
this.latch = latch;
}
public void run() {
System.out.println("thread started: ");
try {
Thread.sleep(3000);
} catch (InterruptedException e) {
e.printStackTrace();
}
latch.countDown();
}
}
public class App {
public static void main(String args[]) {
CountDownLatch latch = new CountDownLatch(3);
ExecutorService executorService = Executors.newFixedThreadPool(3);
for (int i = 0; i < 3; i++) {
executorService.submit(new Processor(latch));
}
try {
latch.await();
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
System.out.println("task completed");
}
}
You need to shut down the executor service. Add this line after the for loop:
executorService.shutdown();
An alternative method that waits for all actively executing tasks to terminate is
executorService.awaitTermination();
You can read more in the ExecutorService Javadoc.
You need to shutdown your executor service properly. Or it will wait for new tasks indefinitely.
Add:
executorService.shutdown();
Before:
System.out.println("task completed");
I think, that the main reason, why it was done like that in java api, is that executorService may receive tasks from multiple threads (other than main), so - why should it stop, when there is no more actions in the main thread? Yes, I believe, it should not.

Java non-blocking IO selector causing channel register to block

I have two threads that I'm dealing with Java NIO for non-blocking sockets. This is what the threads are doing:
Thread 1:
A loop that calls on the select() method of a selector. If any keys are available, they are processed accordingly.
Thread 2:
Occasionally registers a SocketChannel to the selector by calling register().
The problem is, unless the timeout for select() is very small (like around 100ms), the call to register() will block indefinitely. Even though the channel is configured to be nonblocking, and the javadocs state that the Selector object is thread safe (but it's selection keys are not, I know).
So anyone have any ideas on what the issue could be? The application works perfectly if I put everything in one thread. No problems occur then, but I'd really like to have separate threads. Any help is appreciated. I've posted my example code below:
Change the select(1000) to select(100) and it'll work. Leave it as select() or select(1000) and it won't.
import java.io.IOException;
import java.net.DatagramSocket;
import java.net.InetSocketAddress;
import java.nio.ByteBuffer;
import java.nio.channels.DatagramChannel;
import java.nio.channels.SelectionKey;
import java.nio.channels.Selector;
import java.util.Iterator;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
public class UDPSocket
{
private DatagramChannel clientChannel;
private String dstHost;
private int dstPort;
private static Selector recvSelector;
private static volatile boolean initialized;
private static ExecutorService eventQueue = Executors.newSingleThreadExecutor();
public static void init()
{
initialized = true;
try
{
recvSelector = Selector.open();
}
catch (IOException e)
{
System.err.println(e);
}
Thread t = new Thread(new Runnable()
{
#Override
public void run()
{
while(initialized)
{
readData();
Thread.yield();
}
}
});
t.start();
}
public static void shutdown()
{
initialized = false;
}
private static void readData()
{
try
{
int numKeys = recvSelector.select(1000);
if (numKeys > 0)
{
Iterator i = recvSelector.selectedKeys().iterator();
while(i.hasNext())
{
SelectionKey key = i.next();
i.remove();
if (key.isValid() && key.isReadable())
{
DatagramChannel channel = (DatagramChannel) key.channel();
// allocate every time we receive so that it's a copy that won't get erased
final ByteBuffer buffer = ByteBuffer.allocate(Short.MAX_VALUE);
channel.receive(buffer);
buffer.flip();
final SocketSubscriber subscriber = (SocketSubscriber) key.attachment();
// let user handle event on a dedicated thread
eventQueue.execute(new Runnable()
{
#Override
public void run()
{
subscriber.onData(buffer);
}
});
}
}
}
}
catch (IOException e)
{
System.err.println(e);
}
}
public UDPSocket(String dstHost, int dstPort)
{
try
{
this.dstHost = dstHost;
this.dstPort = dstPort;
clientChannel = DatagramChannel.open();
clientChannel.configureBlocking(false);
}
catch (IOException e)
{
System.err.println(e);
}
}
public void addListener(SocketSubscriber subscriber)
{
try
{
DatagramChannel serverChannel = DatagramChannel.open();
serverChannel.configureBlocking(false);
DatagramSocket socket = serverChannel.socket();
socket.bind(new InetSocketAddress(dstPort));
SelectionKey key = serverChannel.register(recvSelector, SelectionKey.OP_READ);
key.attach(subscriber);
}
catch (IOException e)
{
System.err.println(e);
}
}
public void send(ByteBuffer buffer)
{
try
{
clientChannel.send(buffer, new InetSocketAddress(dstHost, dstPort));
}
catch (IOException e)
{
System.err.println(e);
}
}
public void close()
{
try
{
clientChannel.close();
}
catch (IOException e)
{
System.err.println(e);
}
}
}
import java.nio.ByteBuffer;
public interface SocketSubscriber
{
public void onData(ByteBuffer data);
}
Example usage:
public class Test implements SocketSubscriber
{
public static void main(String[] args) throws Exception
{
UDPSocket.init();
UDPSocket test = new UDPSocket("localhost", 1234);
test.addListener(new Test());
UDPSocket test2 = new UDPSocket("localhost", 4321);
test2.addListener(new Test());
System.out.println("Listening...");
ByteBuffer buffer = ByteBuffer.allocate(500);
test.send(buffer);
buffer.rewind();
test2.send(buffer);
System.out.println("Data sent...");
Thread.sleep(5000);
UDPSocket.shutdown();
}
#Override
public void onData(ByteBuffer data)
{
System.out.println("Received " + data.limit() + " bytes of data.");
}
}
The Selector has several documented levels of internal synchronization, and you are running into them all. Call wakeup() on the selector before you call register(). Make sure the select() loop works correctly if there are zero selected keys, which is what will happen on wakeup().
I ran into the same issue today (that is "wakeupAndRegister" not being available). I hope my solution might be helpful:
Create a sync object:
Object registeringSync = new Object();
Register a channel by doing:
synchronized (registeringSync) {
selector.wakeup(); // Wakes up a CURRENT or (important) NEXT select
// !!! Might run into a deadlock "between" these lines if not using the lock !!!
// To force it, insert Thread.sleep(1000); here
channel.register(selector, ...);
}
The thread should do the following:
public void run() {
while (initialized) {
if (selector.select() != 0) { // Blocks until "wakeup"
// Iterate through selected keys
}
synchronized (registeringSync) { } // Cannot continue until "register" is complete
}
}

Categories