Overview  Package   Class  Use  Deprecated  Index
Home >> All >> org >> ematgine >> utils >> [ concurrent overview ]	PREV CLASS  NEXT CLASS
SUMMARY:  JAVADOC |  SOURCE |  DOWNLOAD | NESTED | FIELD | CONSTR | METHOD	DETAIL: FIELD | CONSTR | METHOD

---

org.ematgine.utils.concurrent
Class PooledExecutor  

java.lang.Object
  org.ematgine.utils.concurrent.ThreadFactoryUser
      org.ematgine.utils.concurrent.PooledExecutor

All Implemented Interfaces:

Executor

---

public class PooledExecutor

extends ThreadFactoryUser

implements Executor

A tunable, extensible thread pool class. The main supported public method is execute(Runnable command), which can be called instead of directly creating threads to execute commands.

Thread pools can be useful for several, usually intertwined reasons:

• To bound resource use. A limit can be placed on the maximum number of simultaneously executing threads.
• To manage concurrency levels. A targeted number of threads can be allowed to execute simultaneously.
• To manage a set of threads performing related tasks.
• To minimize overhead, by reusing previously constructed Thread objects rather than creating new ones. (Note however that pools are hardly ever cure-alls for performance problems associated with thread construction, especially on JVMs that themselves internally pool or recycle threads.)

These goals introduce a number of policy parameters that are encapsulated in this class. All of these parameters have defaults and are tunable, either via get/set methods, or, in cases where decisions should hold across lifetimes, via methods that can be easily overridden in subclasses. The main, most commonly set parameters can be established in constructors. Policy choices across these dimensions can and do interact. Be careful. See the usage examples below.

Queueing

By default, this pool uses queueless synchronous channels to to hand off work to threads. This is a safe, conservative policy that avoids lockups when handling sets of requests that might have internal dependencies. (In these cases, queuing one task could lock up another that would be able to continue if the queued task were to run.) If you are sure that this cannot happen, then you can instead supply a queue of some sort (for example, a BoundedBuffer or LinkedQueue) in the constructor. This will cause new commands to be queued in cases where all MaximumPoolSize threads are busy. Queues are sometimes appropriate when each task is completely independent of others, so tasks cannot affect each others execution. For example, in an http server.

When given a choice, this pool always prefers adding a new thread rather than queueing if there are currently fewer than the current getMinimumPoolSize threads running, but otherwise always prefers queuing a request rather than adding a new thread. Thus, if you use an unbounded buffer, you will never have more than getMinimumPoolSize threads running. (Since the default minimumPoolSize is one, you will probably want to explicitly setMinimumPoolSize.)

While queuing can be useful in smoothing out transient bursts of requests, especially in socket-based services, it is not very well behaved when commands continue to arrive on average faster than they can be processed. Using bounds for both the queue and the pool size, along with run-when-blocked policy is often a reasonable response to such possibilities.

Queue sizes and maximum pool sizes can often be traded off for each other. Using large queues and small pools minimizes CPU usage, OS resources, and context-switching overhead, but can lead to artifically low throughput. Especially if tasks frequently block (for example if they are I/O bound), a JVM and underlying OS may be able to schedule time for more threads than you otherwise allow. Use of small queues or queueless handoffs generally requires larger pool sizes, which keeps CPUs busier but may encounter unacceptable scheduling overhead, which also decreases throughput.

Maximum Pool size

The maximum number of threads to use, when needed. The pool does not by default preallocate threads. Instead, a thread is created, if necessary and if there are fewer than the maximum, only when an execute request arrives. The default value is (for all practical purposes) infinite -- Integer.MAX_VALUE, so should be set in the constructor or the set method unless you are just using the pool to minimize construction overhead. Because task handoffs to idle worker threads require synchronization that in turn relies on JVM scheduling policies to ensure progress, it is possible that a new thread will be created even though an existing worker thread has just become idle but has not progressed to the point at which it can accept a new task. This phenomenon tends to occur on some JVMs when bursts of short tasks are executed.

Minimum Pool size

The minimum number of threads to use, when needed (default 1). When a new request is received, and fewer than the minimum number of threads are running, a new thread is always created to handle the request even if other worker threads are idly waiting for work. Otherwise, a new thread is created only if there are fewer than the maximum and the request cannot immediately be queued.

Preallocation

You can override lazy thread construction policies via method createThreads, which establishes a given number of warm threads. Be aware that these preallocated threads will time out and die (and later be replaced with others if needed) if not used within the keep-alive time window. If you use preallocation, you probably want to increase the keepalive time. The difference between setMinimumPoolSize and createThreads is that createThreads immediately establishes threads, while setting the minimum pool size waits until requests arrive.

Keep-alive time

If the pool maintained references to a fixed set of threads in the pool, then it would impede garbage collection of otherwise idle threads. This would defeat the resource-management aspects of pools. One solution would be to use weak references. However, this would impose costly and difficult synchronization issues. Instead, threads are simply allowed to terminate and thus be GCable if they have been idle for the given keep-alive time. The value of this parameter represents a trade-off between GCability and construction time. In most current Java VMs, thread construction and cleanup overhead is on the order of milliseconds. The default keep-alive value is one minute, which means that the time needed to construct and then GC a thread is expended at most once per minute.

To establish worker threads permanently, use a negative argument to setKeepAliveTime.

Blocked execution policy

If the maximum pool size or queue size is bounded, then it is possible for incoming execute requests to block. There are three supported policies for handling this problem, and mechanics (based on the Strategy Object pattern) to allow others in subclasses:

Run (the default)

The thread making the execute request runs the task itself. This policy helps guard against lockup.

Wait

Wait until a thread becomes available.

Discard

Throw away the current request and return.

Other plausible policies include raising the maximum pool size after checking with some other objects that this is OK.

(Again, these cases can never occur if the maximum pool size is unbounded or the queue is unbounded. In these cases you instead face potential resource exhaustion.) The execute method does not throw any checked exceptions in any of these cases since any errors associated with them must normally be dealt with via handlers or callbacks. (Although in some cases, these might be associated with throwing unchecked exceptions.) You may wish to add special implementations even if you choose one of the listed policies. For example, the supplied Discard policy does not inform the caller of the drop. You could add your own version that does so. Since choice of policies is normally a system-wide decision, selecting a policy affects all calls to execute. If for some reason you would instead like to make per-call decisions, you could add variant versions of the execute method (for example, executeIfWouldNotBlock) in subclasses.

Thread construction parameters

A settable ThreadFactory establishes each new thread. By default, it merely generates a new instance of class Thread, but can be changed to use a Thread subclass, to set priorities, ThreadLocals, etc.

Interruption policy

Worker threads check for interruption after processing each command, and terminate upon interruption. Fresh threads will replace them if needed. Thus, new tasks will not start out in an interrupted state due to an uncleared interruption in a previous task. Also, unprocessed commands are never dropped upon interruption. It would conceptually suffice simply to clear interruption between tasks, but implementation characteristics of interruption-based methods are uncertain enough to warrant this conservative strategy. It is a good idea to be equally conservative in your code for the tasks running within pools. Normally, before shutting down a pool via method interruptAll, you should make sure that all clients of the pool are themselves terminated, in order to prevent hanging or lost commands. Additionally, if you are using some form of queuing, you may wish to call method drain() to remove (and return) unprocessed commands from the queue after shutting down the pool and its clients. If you need to be sure these commands are processed, you can then run() each of the commands in the list returned by drain().

Usage examples.

Probably the most common use of pools is in statics or singletons accessible from a number of classes in a package; for example:

 class MyPool {
   // initialize to use a maximum of 8 threads.
   static PooledExecutor pool = new PooledExecutor(8);
 }
 

Here are some sample variants in initialization:

1. Using a bounded buffer of 10 tasks, at least 4 threads (started only when needed due to incoming requests), but allowing up to 100 threads if the buffer gets full.

           pool = new PooledExecutor(new BoundedBuffer(10), 100);
           pool.setMinimumPoolSize(4);
        

2. Same as (1), except pre-start 9 threads, allowing them to die if they are not used for five minutes.

           pool = new PooledExecutor(new BoundedBuffer(10), 100);
           pool.setMinimumPoolSize(4);
           pool.setKeepAliveTime(1000 * 60 * 5);
           pool.createThreads(9);
        

3. Same as (2) except clients block if both the buffer is full and all 100 threads are busy:

           pool = new PooledExecutor(new BoundedBuffer(10), 100);
           pool.setMinimumPoolSize(4);
           pool.setKeepAliveTime(1000 * 60 * 5);
           pool.waitWhenBlocked();
           pool.createThreads(9);
        

4. An unbounded queue serviced by exactly 5 threads:

           pool = new PooledExecutor(new LinkedQueue());
           pool.setKeepAliveTime(-1); // live forever
           pool.createThreads(5);
        

Usage notes.

Pools do not mesh well with using thread-specific storage via java.lang.ThreadLocal. ThreadLocal relies on the identity of a thread executing a particular task. Pools use the same thread to perform different tasks.

If you need a policy not handled by the parameters in this class consider writing a subclass.

Version note: Previous versions of this class relied on ThreadGroups for aggregate control. This has been removed, and the method interruptAll added, to avoid differences in behavior across JVMs.

[ Introduction to this package. ]

---

Nested Class Summary
protected  class	PooledExecutor.BlockedExecutionHandler           Class for actions to take when execute() blocks.
protected  class	PooledExecutor.DiscardWhenBlocked           Class defining Discard action
protected  class	PooledExecutor.RunWhenBlocked           Class defining Run action
protected  class	PooledExecutor.WaitWhenBlocked           Class defining Wait action
protected  class	PooledExecutor.Worker           Class defining the basic run loop for pooled threads.

 

Nested classes inherited from class org.ematgine.utils.concurrent.ThreadFactoryUser

ThreadFactoryUser.DefaultThreadFactory

 

Field Summary
protected  PooledExecutor.BlockedExecutionHandler	blockedExecutionHandler_           The current handler
static long	DEFAULT_KEEPALIVETIME           The maximum time to keep worker threads alive waiting for new tasks; used if not otherwise specified.
static int	DEFAULT_MAXIMUMPOOLSIZE           The maximum pool size; used if not otherwise specified.
static int	DEFAULT_MINIMUMPOOLSIZE           The minimum pool size; used if not otherwise specified.
protected  Channel	handOff_           The channel is used to hand off the command to a thread in the pool
protected  long	keepAliveTime_
protected  int	maximumPoolSize_
protected  int	minimumPoolSize_
protected  java.lang.Object	poolLock_           Lock used for protecting poolSize_ and threads_ map
protected  int	poolSize_           Current pool size.
protected  java.util.Map	threads_           The set of active threads, declared as a map from workers to their threads.

 

Fields inherited from class org.ematgine.utils.concurrent.ThreadFactoryUser

threadFactory_

 

Constructor Summary

PooledExecutor()           Create a new pool with all default settings

PooledExecutor(Channel channel)           Create a new pool that uses the supplied Channel for queuing, and with all default parameter settings.

PooledExecutor(Channel channel, int maxPoolSize)           Create a new pool that uses the supplied Channel for queuing, and with all default parameter settings except for maximum pool size.

PooledExecutor(int maxPoolSize)           Create a new pool with all default settings except for maximum pool size.

 

Method Summary
protected  void	addThread(java.lang.Runnable command)           Create and start a thread to handle a new command.
int	createThreads(int numberOfThreads)           Create and start up to numberOfThreads threads in the pool.
void	discardWhenBlocked()           Set the policy for blocked execution to be to return without executing the request
java.util.List	drain()           Remove all unprocessed tasks from pool queue, and return them in a java.util.List.
void	execute(java.lang.Runnable command)           Arrange for the given command to be executed by a thread in this pool.
protected  PooledExecutor.BlockedExecutionHandler	getBlockedExecutionHandler()           Get the handler for blocked execution
long	getKeepAliveTime()           Return the number of milliseconds to keep threads alive waiting for new commands.
int	getMaximumPoolSize()           Return the maximum number of threads to simultaneously execute New requests will be handled according to the current blocking policy once this limit is exceeded.
int	getMinimumPoolSize()           Return the minimum number of threads to simultaneously execute.
int	getPoolSize()           Return the current number of active threads in the pool.
protected  java.lang.Runnable	getTask()           get a task from the handoff queue
void	interruptAll()           Interrupt all threads in the pool, causing them all to terminate.
void	runWhenBlocked()           Set the policy for blocked execution to be that the current thread executes the command if there are no available threads in the pool.
void	setKeepAliveTime(long msecs)           Set the number of milliseconds to keep threads alive waiting for new commands.
void	setMaximumPoolSize(int newMaximum)           Set the maximum number of threads to use.
void	setMinimumPoolSize(int newMinimum)           Set the minimum number of threads to use.
void	waitWhenBlocked()           Set the policy for blocked execution to be to wait until a thread is available.
protected  void	workerDone(PooledExecutor.Worker w)           Called upon termination of worker thread

 

Methods inherited from class org.ematgine.utils.concurrent.ThreadFactoryUser

getThreadFactory, setThreadFactory

 

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

 

Field Detail

DEFAULT_MAXIMUMPOOLSIZE

public static final int DEFAULT_MAXIMUMPOOLSIZE

The maximum pool size; used if not otherwise specified. Default value is essentially infinite (Integer.MAX_VALUE)

See Also:

Constant Field Values

---

DEFAULT_MINIMUMPOOLSIZE

public static final int DEFAULT_MINIMUMPOOLSIZE

The minimum pool size; used if not otherwise specified. Default value is 1.

See Also:

Constant Field Values

---

DEFAULT_KEEPALIVETIME

public static final long DEFAULT_KEEPALIVETIME

The maximum time to keep worker threads alive waiting for new tasks; used if not otherwise specified. Default value is one minute (60000 milliseconds).

See Also:

Constant Field Values

---

maximumPoolSize_

protected volatile int maximumPoolSize_

---

minimumPoolSize_

protected volatile int minimumPoolSize_

---

keepAliveTime_

protected long keepAliveTime_

---

handOff_

protected final Channel handOff_

The channel is used to hand off the command to a thread in the pool

---

poolLock_

protected java.lang.Object poolLock_

Lock used for protecting poolSize_ and threads_ map

---

poolSize_

protected volatile int poolSize_

Current pool size. Relies on poolLock_ for all locking. But is also volatile to allow simpler checking inside worker thread runloop.

---

threads_

protected final java.util.Map threads_

The set of active threads, declared as a map from workers to their threads. This is needed by the interruptAll method. It may also be useful in subclasses that need to perform other thread management chores. All operations on the Map should be done holding synchronization on poolLock.

---

blockedExecutionHandler_

protected PooledExecutor.BlockedExecutionHandler blockedExecutionHandler_

The current handler

Constructor Detail

PooledExecutor

public PooledExecutor()

Create a new pool with all default settings

---

PooledExecutor

public PooledExecutor(int maxPoolSize)

Create a new pool with all default settings except for maximum pool size.

---

PooledExecutor

public PooledExecutor(Channel channel)

Create a new pool that uses the supplied Channel for queuing, and with all default parameter settings.

---

PooledExecutor

public PooledExecutor(Channel channel,
                      int maxPoolSize)

Create a new pool that uses the supplied Channel for queuing, and with all default parameter settings except for maximum pool size.

Method Detail

getMaximumPoolSize

public int getMaximumPoolSize()

Return the maximum number of threads to simultaneously execute New requests will be handled according to the current blocking policy once this limit is exceeded.

---

setMaximumPoolSize

public void setMaximumPoolSize(int newMaximum)

Set the maximum number of threads to use. Decreasing the pool size will not immediately kill existing threads, but they may later die when idle.

---

getMinimumPoolSize

public int getMinimumPoolSize()

Return the minimum number of threads to simultaneously execute. (Default value is 1). If fewer than the mininum number are running upon reception of a new request, a new thread is started to handle this request.

---

setMinimumPoolSize

public void setMinimumPoolSize(int newMinimum)

Set the minimum number of threads to use.

---

getPoolSize

public int getPoolSize()

Return the current number of active threads in the pool. This number is just a snaphot, and may change immediately upon returning

---

addThread

protected void addThread(java.lang.Runnable command)

Create and start a thread to handle a new command. Call only when holding poolLock.

---

createThreads

public int createThreads(int numberOfThreads)

Create and start up to numberOfThreads threads in the pool. Return the number created. This may be less than the number requested if creating more would exceed maximum pool size bound.

---

interruptAll

public void interruptAll()

Interrupt all threads in the pool, causing them all to terminate. Assuming that executed tasks do not disable (clear) interruptions, each thread will terminate after processing its current task. Threads will terminate sooner if the executed tasks themselves respond to interrupts.

---

drain

public java.util.List drain()

Remove all unprocessed tasks from pool queue, and return them in a java.util.List. It should normally be used only when there are not any active clients of the pool (otherwise you face the possibility that the method will loop pulling out tasks as clients are putting them in.) This method can be useful after shutting down a pool (via interruptAll) to determine whether there are any pending tasks that were not processed. You can then, for example execute all unprocessed commands via code along the lines of:

   List tasks = pool.drain();
   for (Iterator it = tasks.iterator(); it.hasNext();) 
     ( (Runnable)(it.next()) ).run();
 

---

getKeepAliveTime

public long getKeepAliveTime()

Return the number of milliseconds to keep threads alive waiting for new commands. A negative value means to wait forever. A zero value means not to wait at all.

---

setKeepAliveTime

public void setKeepAliveTime(long msecs)

Set the number of milliseconds to keep threads alive waiting for new commands. A negative value means to wait forever. A zero value means not to wait at all.

---

workerDone

protected void workerDone(PooledExecutor.Worker w)

Called upon termination of worker thread

---

getTask

protected java.lang.Runnable getTask()
                              throws java.lang.InterruptedException

get a task from the handoff queue

---

getBlockedExecutionHandler

protected PooledExecutor.BlockedExecutionHandler getBlockedExecutionHandler()

Get the handler for blocked execution

---

runWhenBlocked

public void runWhenBlocked()

Set the policy for blocked execution to be that the current thread executes the command if there are no available threads in the pool.

---

waitWhenBlocked

public void waitWhenBlocked()

Set the policy for blocked execution to be to wait until a thread is available.

---

discardWhenBlocked

public void discardWhenBlocked()

Set the policy for blocked execution to be to return without executing the request

---

execute

public void execute(java.lang.Runnable command)
             throws java.lang.InterruptedException

Arrange for the given command to be executed by a thread in this pool. The method normally returns when the command has been handed off for (possibly later) execution.

Specified by:

execute in interface Executor

---

Overview  Package   Class  Use  Deprecated  Index
Home >> All >> org >> ematgine >> utils >> [ concurrent overview ]	PREV CLASS  NEXT CLASS
SUMMARY:  JAVADOC |  SOURCE |  DOWNLOAD | NESTED | FIELD | CONSTR | METHOD	DETAIL: FIELD | CONSTR | METHOD

---