Multithreading - Schedule neverending runnables in Threadpools - java

I cannot seem to find a good answer to my problem:
Suppose I have a certain amount of runnables of the form
public class NeverendingRunner implements Runnable {
//...
#Override
public void run() {
while(true){
//Do what you have to do
//At the end of the iteration, get back in queue
}
}
If I use thread pools to run these, they will never left the execution, obviously.
I'd like to schedule them like in the comment I've made at the end of the iteration.
E.G.: I have 2 Runnables with infinite loops (and they have to be infinite for the task they do) and a treadpool of 1 thread (for simplicity). I want these two loops to alternate the usage of the thread. The Runnables don't know each other.
I don't know if this is even possible to be done.
Thank you for any help or suggest
EDIT: Jost solution solved the problem offering the scalability I was aiming for. Thank you.

I think you should use a ThreadPoolExecutor, remove the infinite loop and enable the two Runnables to reschedule themselves to the executor. This code snippet might help:
public class Task implements Runnable
{
private ReentrantLock lock;
private ExecutorService executor;
public Task(ExecutorService executor)
{
this.executor=executor;
}
#Override
public void run()
{
//do some stuff
//...
lock.lock();
executor.execute(this);
lock.unlock();
}
}
Note: I have not tried this example - but in general it should work. The lock ensures that the executor doesn't start with run() before it was finished (sounds a bit weird - but makes sense in the case of more than one thread in the pool).
*Jost

I can imagine something like this
final Runnable task1 = new Runnable() {
public void run() {
for(int i = 0; i < 10; i++) {
System.out.println("task1");
}
}
};
final Runnable task2 = new Runnable() {
public void run() {
for(int i = 0; i < 10; i++) {
System.out.println("task2");
}
}
};
new Thread() {
public void run() {
for(;;) {
task1.run();
task2.run();
}
};
}.start();

The below code may achieve this but its not the best implementation
CyclicBarrier run1 = new CyclicBarrier(2);
CyclicBarrier run2 = new CyclicBarrier(2);
public class RUN1 implements Runnable {
#Override
public void run() {
while(true){
//Do what you have to do
//At the end of the iteration, get back in queue
run1.await();
run2.await();
}
}
}
public class RUN2 implements Runnable {
#Override
public void run() {
while(true){
run1.await();
//Do what you have to do
//At the end of the iteration, get back in queue
run2.await();
}
}
}

If the runnables do not knwo each other, they probably don't have to be run by the same thread. In any case, it is often better to think in terms of runnables and their possible dependencies than in threads. If there is a real reason for them to run alternating (maybe even alternating at defined switch-points, then perhaps they (or what happens inside them) should know of each other.
Normally, when running independent runnables in independent threads, the OS scheduling takes care of alternating them already. Since they are neverending, make sure you wrap the loops in try/catch(InterruptedException) blocks to be able to kill the threads nicely.

See method loopAct() in the file DataflowVariable. It is overhead for your case, but idea is simple: keep a boolean variable showing if the task is working; when the task finishes all jobs, it turns off the variable and exits; when another thread supplies a job to the task, and the variable shows it is not working, it turns on the variable and sends the task to the executor. This way you'll never get the task executed by more than one thread simultaneously.

Related

What is the cleanest way to create, start, and manage long-running threads?

Threads add a lot of verbal to the code and make it harder to understand and reason about. Look at this code for example:
public class ConnectionListener implements Runnable {
private Thread thread;
private boolean running;
public void start() {
if (!running) {
thread = new Thread(this);
thread.start();
}
}
public void stop() {
if (running) {
running = false;
thread.interrupt();
}
}
#Override
public void run() {
running = true;
while (running) {
// Do some crap
}
}
}
The whole concern of this class should be listening for connection requests from the network. But look how many lines of code are added just for managing a thread. Is there any way to make this code cleaner?! I don't want to see the thread = new Thread();, not the thread variable and not any of the stop()/start() methods!
Of course I know what the ExecutorService is... But what if I want to manage a long-running thread? By long-running thread, I mean a thread with a life cycle long as the application's life cycle.
Do you have any good solution for me? A way to remove the thread creation and management concerns from a class without making the class extend another class?
I solved the problem by using a single-threaded executor service. I've also read about the performance differences between Plain Thread, ThreadPool and SingleThreadExecutor - SingleThreadExecutor VS plain thread.
Using a single thread executor allows me to start a single thread and manage it using its Future. See code example:
public void func(String[] args) {
ExecutorService es = Executors.newSingleThreadExecutor();
Future<?> f = es.submit(Some Runnable);
}
Thanks to #BasilBourque that gave me this solution in the comments.

Waking up a thread without risking to get blocked

I have a worker thread running indefinitely, which goes to sleep for one minute if there's nothing to do. Sometimes, another piece of code produces some work and wants to wake the worker thread immediately.
So I did something like this (code for illustration only):
class Worker {
public void run() {
while (!shuttingDown()) {
step();
}
}
private synchronized void step() {
if (hasWork()) {
doIt();
} else {
wait(60_000);
}
}
public synchronized wakeMeUpInside() {
notify();
}
}
What I dislike is having to enter the monitor only for waking something up, which means that the notifying thread may be delayed for no good reason. As the choices of native synchronization are limited, I thought I'd switch to Condition, but it has exactly the same problem:
An implementation may (and typically does) require that the current thread hold the lock associated with this Condition when this method is called.
Here's a semaphore based solution:
class Worker {
// If 0 there's no work available
private workAvailableSem = new Semaphore(0);
public void run() {
while (!shuttingDown()) {
step();
}
}
private synchronized void step() {
// Try to obtain a permit waiting up to 60 seconds to get one
boolean hasWork = workAvailableSem.tryAquire(1, TimeUnit.MINUTES);
if (hasWork) {
doIt();
}
}
public wakeMeUpInside() {
workAvailableSem.release(1);
}
}
I'm not 100% sure this meets your needs. A few things to note:
This will add one permit each time wakeMeUpInside is called. Thus if two threads wake up the Worker it will run doIt twice without blocking. You can extend the example to avoid that.
This waits 60 seconds for work to do. If none is available it'll end up back in the run method which will send it immediately back to the step method which will just wait again. I did this because I'm assuming you had some reason why you wanted to run every 60 seconds even if there's no work. If that's not the case just call aquire and you'll wait indefinitely for work.
As per comments below the OP wants to run only once. While you could call drainPermits in that case a cleaner solution is just to use a LockSupport like so:
class Worker {
// We need a reference to the thread to wake it
private Thread workerThread = null;
// Is there work available
AtomicBoolean workAvailable = new AtomicBoolean(false);
public void run() {
workerThread = Thread.currentThread();
while (!shuttingDown()) {
step();
}
}
private synchronized void step() {
// Wait until work is available or 60 seconds have passed
ThreadSupport.parkNanos(TimeUnit.MINUTES.toNanos(1));
if (workAvailable.getAndSet(false)) {
doIt();
}
}
public wakeMeUpInside() {
// NOTE: potential race here depending on desired semantics.
// For example, if doIt() will do all work we don't want to
// set workAvailable to true if the doIt loop is running.
// There are ways to work around this but the desired
// semantics need to be specified.
workAvailable.set(true);
ThreadSupport.unpark(workerThread);
}
}

Threading Wait During While Loop

My application has 1 global driver, which is responsible for doing the low-level work.
I then have 2 threads, both of which use infinite loops to get some work done. My question is how to allow 1 thread to use the driver as much as possible, but giving a chance to the second thread to use it when necessary.
To elaborate, the code I have is as follows:
public class Game {
private static final Object LOCK = new Object();
private static final Logger LOGGER = Logger.getLogger(Game.class);
private WebDriverController controller;
public Game(WebDriverController controler) {
this.controller = controller;
}
public void startThreadA() {
new Thread(new Runnable() {
#Override
public void run() {
while (true) {
synchronized (LOCK) {
controller.doSomethingA();
}
}
}
}).start();
}
public void startThreadB() {
new Thread(new Runnable() {
#Override
public void run() {
while (true) {
...
...
synchronized (LOCK) {
controller.doSomethingB();
}
...
...
}
}
}).start();
}
}
The logic is to allow the first thread to execute doSomethingA() as much as possible, with the second thread only acquiring the lock to complete little tasks and then giving the lock back to the first thread.
Using this code, the first thread will continuously use the controller to do what it needs to do, whereas the second thread gets stuck waiting at its synchronized block. The way I have currently fixed this is by adding a pause to the first thread, to give the second thread a chance to acquire the lock, as follows:
public void startThreadA() {
new Thread(new Runnable() {
#Override
public void run() {
while (true) {
synchronized (LOCK) {
controller.doSomethingA();
}
try {
Thread.sleep(1);
} catch (InterruptedException e) {
LOGGER.error(null, e);
}
}
}
}).start();
}
This does work exactly as intended, but it doesn't seem right. I'm not happy with the manual pause after each iteration, especially if the second thread does not need the lock as it's wasting time.
What do I replace the pause with to make this more efficient?
Why you use synchronized in run()? Use synchronized or Lock in your methods in WebDriverController.
public void doSomeThingA(){
lock.lock();
try {
//your stuff
} finally {
lock.unlock();
}
}
And in run method of Thread invoke these methods.
I think you are approaching this from the wrong direction, as in your current setup 99.999% of the time thread A calls for a monitor the processing time is wasted. However as I do not have enough details about your actual problem, here is a quick solution using a ReentrantLock with fair scheduling (FIFO):
protected final ReentrantLock lock = new ReentrantLock(true); // fair scheduling
public void functionA() {
lock.lock();
try {
controller.functionA();
} finally {
lock.unlock();
}
}
public void functionB() {
lock.lock();
try {
controller.functionB();
} finally {
lock.unlock();
}
}
Explanation:
If Thread A is currently holding the lock and Thread B calls, B is guaranteed to receive the monitor right after A releases it, even if A immediately (before any thread switch occurs) calls for it again.
There are a few options here. The best bet in this instance is likely to be remove the responsibility of deciding when to do work from the threads and instead, waiting for an event from a monitor to release the threads to do work. You can then schedule the work in whichever ratio is best suited to the purpose.
Alternatively, remove the lack of thread safety from your controller code.
Assuming that above thread organization is the best way to go for your particular case, your problem is that first thread holds the lock too long, thus starving the second one.
You can check if doSomethingA function really needs locked driver all the time while it is being executed (in most cases it doesn't), and if not split it into multiple smaller execution blocks, some of which hold the lock while other's don't. This will create more time for second thread to kick in when it needs to.
If that cannot be done then you really need to rethink your app, because you have created a resource bottleneck.
It looks like Thread.yield () is what you are looking for.

Java threads scheduling

I have a MainClass, a Worker class and a Supervisor class. In MainClass i create 10 Worker classes and a Supervisor class that run in separate threads.
class MainClass {
public static void main(String args[]) {
for (int i=0; i<10 ;i++) {
Thread t = new Thread( new Worker());
t.start();
}
(new Thread(new Supervisor()).start();
}
.
class Worker extends Thread {
public void run() {
while(true) {
if(some_condition) {
//do stuff
} else {
// pause thread execution for undefined time.
}
}
}
}
.
class Supervisor extends Thread {
public void run() {
while(true) {
if(some_condition) {
// restart Workers thread that are paused.
}
// do other stuff
}
}
}
I don't know how to implement this, cause the conditions in every thread are independent from each other so i don't need to synchronize, so i can't use wait-notify.
I don't know how to implement this, cause the conditions in every thread are independent from each other so i don't need to synchronize, so i can't use wait-notify.
Sure you can.
The subtlety here is that presumably Supervisor doesn't actually know whether worker threads are really paused. (If it does, then the conditions are not independent.)
Since Supervisor doesn't know whether the threads are actually paused (by assumption), you have to design what you want to happen if it tries to unpause an already-unpaused thread.
a) Should an unpause do nothing?
b) Or should it immediately unpause the next time a worker tries to pause itself?
If the answer is (b), then you have to worry about thread safety. If the answer is (a), then you don't (unless you have some other data to pass between threads!)
Either way, you can still use wait and notify.
As per my understanding u want to create separate thread pools which consist of 10 workers or number as per your requirement.
As far as pools are concerned you can check for ThreadPoolExecutor in java.util.concurrent api. Internally ThreadPoolexecutor also creates worker Threads for running tasks.
Try reading ThreadPoolExecutor it might help you or please elaborate your question whats your ultimate objective you wish to achieve by this problem.

How to know if other threads have finished?

I have an object with a method named StartDownload(), that starts three threads.
How do I get a notification when each thread has finished executing?
Is there a way to know if one (or all) of the thread is finished or is still executing?
There are a number of ways you can do this:
Use Thread.join() in your main thread to wait in a blocking fashion for each Thread to complete, or
Check Thread.isAlive() in a polling fashion -- generally discouraged -- to wait until each Thread has completed, or
Unorthodox, for each Thread in question, call setUncaughtExceptionHandler to call a method in your object, and program each Thread to throw an uncaught Exception when it completes, or
Use locks or synchronizers or mechanisms from java.util.concurrent, or
More orthodox, create a listener in your main Thread, and then program each of your Threads to tell the listener that they have completed.
How to implement Idea #5? Well, one way is to first create an interface:
public interface ThreadCompleteListener {
void notifyOfThreadComplete(final Thread thread);
}
then create the following class:
public abstract class NotifyingThread extends Thread {
private final Set<ThreadCompleteListener> listeners
= new CopyOnWriteArraySet<ThreadCompleteListener>();
public final void addListener(final ThreadCompleteListener listener) {
listeners.add(listener);
}
public final void removeListener(final ThreadCompleteListener listener) {
listeners.remove(listener);
}
private final void notifyListeners() {
for (ThreadCompleteListener listener : listeners) {
listener.notifyOfThreadComplete(this);
}
}
#Override
public final void run() {
try {
doRun();
} finally {
notifyListeners();
}
}
public abstract void doRun();
}
and then each of your Threads will extend NotifyingThread and instead of implementing run() it will implement doRun(). Thus when they complete, they will automatically notify anyone waiting for notification.
Finally, in your main class -- the one that starts all the Threads (or at least the object waiting for notification) -- modify that class to implement ThreadCompleteListener and immediately after creating each Thread add itself to the list of listeners:
NotifyingThread thread1 = new OneOfYourThreads();
thread1.addListener(this); // add ourselves as a listener
thread1.start(); // Start the Thread
then, as each Thread exits, your notifyOfThreadComplete method will be invoked with the Thread instance that just completed (or crashed).
Note that better would be to implements Runnable rather than extends Thread for NotifyingThread as extending Thread is usually discouraged in new code. But I'm coding to your question. If you change the NotifyingThread class to implement Runnable then you have to change some of your code that manages Threads, which is pretty straightforward to do.
Solution using CyclicBarrier
public class Downloader {
private CyclicBarrier barrier;
private final static int NUMBER_OF_DOWNLOADING_THREADS;
private DownloadingThread extends Thread {
private final String url;
public DownloadingThread(String url) {
super();
this.url = url;
}
#Override
public void run() {
barrier.await(); // label1
download(url);
barrier.await(); // label2
}
}
public void startDownload() {
// plus one for the main thread of execution
barrier = new CyclicBarrier(NUMBER_OF_DOWNLOADING_THREADS + 1); // label0
for (int i = 0; i < NUMBER_OF_DOWNLOADING_THREADS; i++) {
new DownloadingThread("http://www.flickr.com/someUser/pic" + i + ".jpg").start();
}
barrier.await(); // label3
displayMessage("Please wait...");
barrier.await(); // label4
displayMessage("Finished");
}
}
label0 - cyclic barrier is created with number of parties equal to the number of executing threads plus one for the main thread of execution (in which startDownload() is being executed)
label 1 - n-th DownloadingThread enters the waiting room
label 3 - NUMBER_OF_DOWNLOADING_THREADS have entered the waiting room. Main thread of execution releases them to start doing their downloading jobs in more or less the same time
label 4 - main thread of execution enters the waiting room. This is the 'trickiest' part of the code to understand. It doesn't matter which thread will enter the waiting room for the second time. It is important that whatever thread enters the room last ensures that all the other downloading threads have finished their downloading jobs.
label 2 - n-th DownloadingThread has finished its downloading job and enters the waiting room. If it is the last one i.e. already NUMBER_OF_DOWNLOADING_THREADS have entered it, including the main thread of execution, main thread will continue its execution only when all the other threads have finished downloading.
You should really prefer a solution that uses java.util.concurrent. Find and read Josh Bloch and/or Brian Goetz on the topic.
If you are not using java.util.concurrent.* and are taking responsibility for using Threads directly, then you should probably use join() to know when a thread is done. Here is a super simple Callback mechanism. First extend the Runnable interface to have a callback:
public interface CallbackRunnable extends Runnable {
public void callback();
}
Then make an Executor that will execute your runnable and call you back when it is done.
public class CallbackExecutor implements Executor {
#Override
public void execute(final Runnable r) {
final Thread runner = new Thread(r);
runner.start();
if ( r instanceof CallbackRunnable ) {
// create a thread to perform the callback
Thread callerbacker = new Thread(new Runnable() {
#Override
public void run() {
try {
// block until the running thread is done
runner.join();
((CallbackRunnable)r).callback();
}
catch ( InterruptedException e ) {
// someone doesn't want us running. ok, maybe we give up.
}
}
});
callerbacker.start();
}
}
}
The other sort-of obvious thing to add to your CallbackRunnable interface is a means to handle any exceptions, so maybe put a public void uncaughtException(Throwable e); line in there and in your executor, install a Thread.UncaughtExceptionHandler to send you to that interface method.
But doing all that really starts to smell like java.util.concurrent.Callable. You should really look at using java.util.concurrent if your project permits it.
Many things have been changed in last 6 years on multi-threading front.
Instead of using join() and lock API, you can use
1.ExecutorService invokeAll() API
Executes the given tasks, returning a list of Futures holding their status and results when all complete.
2.CountDownLatch
A synchronization aid that allows one or more threads to wait until a set of operations being performed in other threads completes.
A CountDownLatch is initialized with a given count. The await methods block until the current count reaches zero due to invocations of the countDown() method, after which all waiting threads are released and any subsequent invocations of await return immediately. This is a one-shot phenomenon -- the count cannot be reset. If you need a version that resets the count, consider using a CyclicBarrier.
3.ForkJoinPool or newWorkStealingPool() in Executors is other way
4.Iterate through all Future tasks from submit on ExecutorService and check the status with blocking call get() on Future object
Have a look at related SE questions:
How to wait for a thread that spawns it's own thread?
Executors: How to synchronously wait until all tasks have finished if tasks are created recursively?
Do you want to wait for them to finish? If so, use the Join method.
There is also the isAlive property if you just want to check it.
You can interrogate the thread instance with getState() which returns an instance of Thread.State enumeration with one of the following values:
* NEW
A thread that has not yet started is in this state.
* RUNNABLE
A thread executing in the Java virtual machine is in this state.
* BLOCKED
A thread that is blocked waiting for a monitor lock is in this state.
* WAITING
A thread that is waiting indefinitely for another thread to perform a particular action is in this state.
* TIMED_WAITING
A thread that is waiting for another thread to perform an action for up to a specified waiting time is in this state.
* TERMINATED
A thread that has exited is in this state.
However I think it would be a better design to have a master thread which waits for the 3 children to finish, the master would then continue execution when the other 3 have finished.
You could also use the Executors object to create an ExecutorService thread pool. Then use the invokeAll method to run each of your threads and retrieve Futures. This will block until all have finished execution. Your other option would be to execute each one using the pool and then call awaitTermination to block until the pool is finished executing. Just be sure to call shutdown() when you're done adding tasks.
I would suggest looking at the javadoc for Thread class.
You have multiple mechanisms for thread manipulation.
Your main thread could join() the three threads serially, and would then not proceed until all three are done.
Poll the thread state of the spawned threads at intervals.
Put all of the spawned threads into a separate ThreadGroup and poll the activeCount() on the ThreadGroup and wait for it to get to 0.
Setup a custom callback or listener type of interface for inter-thread communication.
I'm sure there are plenty of other ways I'm still missing.
I guess the easiest way is to use ThreadPoolExecutor class.
It has a queue and you can set how many threads should be working in parallel.
It has nice callback methods:
Hook methods
This class provides protected overridable beforeExecute(java.lang.Thread, java.lang.Runnable) and afterExecute(java.lang.Runnable, java.lang.Throwable) methods that are called before and after execution of each task. These can be used to manipulate the execution environment; for example, reinitializing ThreadLocals, gathering statistics, or adding log entries. Additionally, method terminated() can be overridden to perform any special processing that needs to be done once the Executor has fully terminated.
which is exactly what we need. We will override afterExecute() to get callbacks after each thread is done and will override terminated() to know when all threads are done.
So here is what you should do
Create an executor:
private ThreadPoolExecutor executor;
private int NUMBER_OF_CORES = Runtime.getRuntime().availableProcessors();
private void initExecutor() {
executor = new ThreadPoolExecutor(
NUMBER_OF_CORES * 2, //core pool size
NUMBER_OF_CORES * 2, //max pool size
60L, //keep aive time
TimeUnit.SECONDS,
new LinkedBlockingQueue<Runnable>()
) {
#Override
protected void afterExecute(Runnable r, Throwable t) {
super.afterExecute(r, t);
//Yet another thread is finished:
informUiAboutProgress(executor.getCompletedTaskCount(), listOfUrisToProcess.size());
}
}
};
#Override
protected void terminated() {
super.terminated();
informUiThatWeAreDone();
}
}
And start your threads:
private void startTheWork(){
for (Uri uri : listOfUrisToProcess) {
executor.execute(new Runnable() {
#Override
public void run() {
doSomeHeavyWork(uri);
}
});
}
executor.shutdown(); //call it when you won't add jobs anymore
}
Inside method informUiThatWeAreDone(); do whatever you need to do when all threads are done, for example, update UI.
NOTE: Don't forget about using synchronized methods since you do your work in parallel and BE VERY CAUTIOUS if you decide to call synchronized method from another synchronized method! This often leads to deadlocks
Hope this helps!
Here's a solution that is simple, short, easy to understand, and works perfectly for me. I needed to draw to the screen when another thread ends; but couldn't because the main thread has control of the screen. So:
(1) I created the global variable: boolean end1 = false; The thread sets it to true when ending. That is picked up in the mainthread by "postDelayed" loop, where it is responded to.
(2) My thread contains:
void myThread() {
end1 = false;
new CountDownTimer(((60000, 1000) { // milliseconds for onFinish, onTick
public void onFinish()
{
// do stuff here once at end of time.
end1 = true; // signal that the thread has ended.
}
public void onTick(long millisUntilFinished)
{
// do stuff here repeatedly.
}
}.start();
}
(3) Fortunately, "postDelayed" runs in the main thread, so that's where in check the other thread once each second. When the other thread ends, this can begin whatever we want to do next.
Handler h1 = new Handler();
private void checkThread() {
h1.postDelayed(new Runnable() {
public void run() {
if (end1)
// resond to the second thread ending here.
else
h1.postDelayed(this, 1000);
}
}, 1000);
}
(4) Finally, start the whole thing running somewhere in your code by calling:
void startThread()
{
myThread();
checkThread();
}
You could also use SwingWorker, which has built-in property change support. See addPropertyChangeListener() or the get() method for a state change listener example.
Look at the Java documentation for the Thread class. You can check the thread's state. If you put the three threads in member variables, then all three threads can read each other's states.
You have to be a bit careful, though, because you can cause race conditions between the threads. Just try to avoid complicated logic based on the state of the other threads. Definitely avoid multiple threads writing to the same variables.

Categories