Getting a memory leak on ScheduledExecutorService

[Tom Davis Sr]

I am trying the basic example from the java ScheduledExecutorService API. After a few minutes, I see the memory usage go up. My task has to run for a very long time, hence I am concerned that it could crash the application. Any input would be appreciated.

import static java.util.concurrent.TimeUnit.SECONDS;

import java.util.concurrent.Executors;
import java.util.concurrent.ScheduledExecutorService;
import java.util.concurrent.ScheduledFuture;
 class BeeperControl {
	 
	  public static void main(String[] args)
		{
		  BeeperControl bctl = new BeeperControl();
		  bctl.beepForAnHour();
		  
		}
	 
    private final ScheduledExecutorService scheduler =
       Executors.newScheduledThreadPool(1);

    public void beepForAnHour() {
        final Runnable beeper = new Runnable() {
                public void run() { System.out.println("beep"); }
            };
        final ScheduledFuture<?> beeperHandle =
            scheduler.scheduleAtFixedRate(beeper, 5, 2, SECONDS);
        scheduler.schedule(new Runnable() {
                public void run() { beeperHandle.cancel(true); }
            }, 60 * 60, SECONDS);
    }
 }

[Nitesh Kant]

It will be better to profile the JVM and find out what is taking memory, otherwise the memory leak is merely a speculation.
You can use jmap at the minimum to get a dump of the heap and later view it using jhat

[Tom Davis Sr]

Thank you Nitesh. I had this sample program (which really has one print statement to execute every 2 secs) running for 2 hours and I noticed that the memory usage on Windows Task Manager is still going up. I expected it to remain constant after a while.

Is it possible that the JVM would still need more memory after all this time?

[Chris Hurst]

JVisualvm in the bin folder of your jdk is also quite useful, you'll get a "saw tooth" graph as memory rises and if that grapgh has an underlying trend upwards you have an issue.

Just leave it running a long time then look at which objects are using the most memory (string's are always near the top but check what those strings are content wise ... I once found 100Mb's of log file in memory from a Sun bug ..lol) then look for any objects you weren't expecting ;-)