I have code where I schedule a task using java.util.Timer
. I was looking around and saw ExecutorService
can do the same. So this question here, hav
According to Java Concurrency in Practice:
Timer
can be sensitive to changes in the system clock, ScheduledThreadPoolExecutor
isn't.Timer
has only one execution thread, so long-running task can delay other tasks. ScheduledThreadPoolExecutor
can be configured with any number of threads. Furthermore, you have full control over created threads, if you want (by providing ThreadFactory
).TimerTask
kill that one thread, thus making Timer
dead :-( ... i.e. scheduled tasks will not run anymore. ScheduledThreadExecutor
not only catches runtime exceptions, but it lets you handle them if you want (by overriding afterExecute
method from ThreadPoolExecutor
). Task which threw exception will be canceled, but other tasks will continue to run.If you can use ScheduledThreadExecutor
instead of Timer
, do so.
One more thing... while ScheduledThreadExecutor
isn't available in Java 1.4 library, there is a Backport of JSR 166 (java.util.concurrent) to Java 1.2, 1.3, 1.4, which has the ScheduledThreadExecutor
class.
Here's some more good practices around Timer use:
http://tech.puredanger.com/2008/09/22/timer-rules/
In general, I'd use Timer for quick and dirty stuff and Executor for more robust usage.
If it's available to you, then it's difficult to think of a reason not to use the Java 5 executor framework. Calling:
ScheduledExecutorService ex = Executors.newSingleThreadScheduledExecutor();
will give you a ScheduledExecutorService
with similar functionality to Timer
(i.e. it will be single-threaded) but whose access may be slightly more scalable (under the hood, it uses concurrent structures rather than complete synchronization as with the Timer
class). Using a ScheduledExecutorService
also gives you advantages such as:
newScheduledThreadPoolExecutor()
or the ScheduledThreadPoolExecutor
class)About the only reasons for sticking to Timer
I can think of are:
My reason for sometimes preferring Timer over Executors.newSingleThreadScheduledExecutor() is that I get much cleaner code when I need the timer to execute on daemon threads.
compare
private final ThreadFactory threadFactory = new ThreadFactory() {
public Thread newThread(Runnable r) {
Thread t = new Thread(r);
t.setDaemon(true);
return t;
}
};
private final ScheduledExecutorService timer = Executors.newSingleThreadScheduledExecutor(threadFactory);
with
private final Timer timer = new Timer(true);
I do this when I don't need the robustness of an executorservice.
ExecutorService is newer and more general. A timer is just a thread that periodically runs stuff you have scheduled for it.
An ExecutorService may be a thread pool, or even spread out across other systems in a cluster and do things like one-off batch execution, etc...
Just look at what each offers to decide.
From Oracle documentation page on ScheduledThreadPoolExecutor
A ThreadPoolExecutor that can additionally schedule commands to run after a given delay, or to execute periodically. This class is preferable to Timer when multiple worker threads are needed, or when the additional flexibility or capabilities of ThreadPoolExecutor (which this class extends) are required.
ExecutorService/ThreadPoolExecutor
or ScheduledThreadPoolExecutor
is obvious choice when you have multiple worker threads.
Pros of ExecutorService
over Timer
Timer
can't take advantage of available CPU cores unlike ExecutorService
especially with multiple tasks using flavours of ExecutorService
like ForkJoinPoolExecutorService
provides collaborative API if you need coordination between multiple tasks. Assume that you have to submit N number of worker tasks and wait for completion of all of them. You can easily achieve it with invokeAll API. If you want to achieve the same with multiple Timer
tasks, it would be not simple. ThreadPoolExecutor provides better API for management of Thread life cycle.
Thread pools address two different problems: they usually provide improved performance when executing large numbers of asynchronous tasks, due to reduced per-task invocation overhead, and they provide a means of bounding and managing the resources, including threads, consumed when executing a collection of tasks. Each ThreadPoolExecutor also maintains some basic statistics, such as the number of completed tasks
Few advantages:
a. You can create/manage/control life cycle of Threads & optimize thread creation cost overheads
b. You can control processing of tasks ( Work Stealing, ForkJoinPool, invokeAll) etc.
c. You can monitor the progress and health of threads
d. Provides better exception handling mechanism