Java Thread Handling

Hi,
I've been poking around in the site but haven't been able to find the answer: does Java dispatch threads to different cores when it has control of the processor(s) or does the JVM internally decide which thread gets control of the processor(s)?
Thanks.

Thanks...I seemed to remember that Java decided but I wasn't sure and was in fact correct.
Speaking of which - can you perhaps link me or tell me what terms to search under so I can get some Sun documentation on this and read up about it? I am most interested in the details.
Thanks !

Similar Messages

Java threads monitoring

I am using Java threads in my Java program, running on Windows Professional 2000. I am monitoring the threads using Windows Performance Monitor (WPM). However, even running a small Java program which creates only 2 threads, results in many (at least 10) instances of Java threads being reported by the WPM. Problem is, in WPM, there is no way to determine which threads in the WPM are the threads that I programmatically created. Any ideas on how to determine this?
Thanks.
Rhodie

Hi Dmirty ,
To handle large asynchronous message in queues you can use message packaging where multiple message are processed in one package . to make it applicable you have to perform these simple steps
1. go to SXMB_ADM add one RUNTIME parameter PACKAGING and value 1
2 GO TO transaction SXMS_BCONF set delay time 0 message count 100 messages and package size 1000 KB
with these setting your improvement will increase .secondly also put IN SXMB_ADM monitor category parameter QRFC_RESTART_ALLOWED TO 1
this will automatically start your queue . Only thing you have to take care if you enable packaging them don't keep too many parallel queue i.e EO_INBOUND_PARALLEL and EO_OUTBOUND_PARALLEL should be less than 20
Regards,
Saurabh

How to terminate a java thread from c++ code?

Hi,
I made a screensaver which loads a jvm, forks a thread running a java class, wait until user's action(i.e. mouse move/click keyboard input), then terminate the java thread.
However, I met a problem, How to terminate a running java thread from c++ code?
Here is my code, but it does not work: (even after the terminate is called, jvm throws an error)
JNIEnv* env;
JavaVM* jvm;
HANDLE hThread; //handle for the startThread
unsigned __stdcall startThread(void *arg)
     jclass cls;
     jmethodID mainId;
     jint          res;
     int threadNum = (int)arg;
     res = jvm->AttachCurrentThread((void**)&env, NULL);
cls = env->FindClass( MAIN_CLASS);
     mainId = env->GetStaticMethodID(cls, "main", "([Ljava/lang/String;)V");
     // setup the parameters to pass to main()
     jstring str;
     jobjectArray args;
     int i=0;
     args = env->NewObjectArray(1, env->FindClass("java/lang/String"), 0); //only one input parameters
     str = env->NewStringUTF("localhost");
     env->SetObjectArrayElement(args, 0, str);
     env->CallStaticVoidMethod(cls, mainId, args); // call main()
     if (env->ExceptionOccurred()) {
               env->ExceptionDescribe();
     return TRUE;
Here is the main method:
First create the jvm and load the thread. then tries to terminate the thread, but failed here
switch (msg)
{ case WM_CREATE:
          JavaVMOption options[NUMBEROFOPTIONS];
          JavaVMInitArgs vmargs;
          jint rc;
          vmargs.version = JNI_VERSION_1_4; /* version 1.4 */
          vmargs.options = options;
          vmargs.nOptions = NUMBEROFOPTIONS;
          vmargs.ignoreUnrecognized = JNI_FALSE;
          rc=JNI_CreateJavaVM( &jvm, (void **)&env, &vmargs ); /* create JVM */
          /* We pass the thread number as the argument to the invoked thread */
          unsigned int threadId = -1;
          // to initialize a thread-safe C runtime library
          hThread = (HANDLE)_beginthreadex(NULL, 0, &startThread, NULL, 0, &threadId );
} break;
case (WM_DESTROY):
          CloseHandle( hThread );
          jvm->DestroyJavaVM(); /* kill JVM */
          Debug("Destroy Java Virtual Machine\n");
}break;
Note, after the thread "startThread" runs, it has an infinite loop and will not terminate, until the user clicks to terminate.(I didn't call jvm->DetachCurrentThread();_endthreadex(0); in the "startThread" because the java thread will not terminate by itself)
Thus, the only way to terminate the thread is to call closehandle(hthread) in the main thread, which may cause jvm to throw an error.
Do you know how to terminate a java thread safely???
Thanks a lot

Assuming that your java thread is in a loop of some kind. Such as
int i=1; /* I tried using a boolean, I just could not get my C++ env, to change this value. So i decided to use an int */
run {
while(i)
isdfjsdfj
void seti()
i=0
So, B/4, i call destroyVM in my C++ code, i call this seti(). so the loop terminates, therefore my thread finishes executing and ends.
Hope this helps
tola.

Help calling JNI function from a java thread

I am using the following code to call a JNI fuction from a java thread:
public class CMSMessageHandler extends MessageHandler{
static{
System.loadLibrary("jastb");
public native int cmsReadIt(ByteBuffer buff);
private ByteBuffer dirbuf = null;
/** Creates a new instance of CMSMessageHandler */
public CMSMessageHandler() {
// allocate to max size datagram
dirbuf = ByteBuffer.allocateDirect(65536);
public void readMessages(){
/*Thread worker = new Thread(){
public void run(){*/
int count;
while (true){
try{
count = cmsReadIt(dirbuf);
} catch (Exception e){
e.printStackTrace();
worker.start();*/
public static void main(String[] args) {
CMSMessageHandler handler = new CMSMessageHandler();
handler.readMessages();
When I run this main method with the thread commented out, it works great. When I run the main method with the thread, it works great for a while, and then funny things start happening with the native library I am using. My native library uses shared memory, signals, and semaphores. It feels a bit like my java program is stomping on the resources that my native library is using, when a run the java program with a thread.
Is there something I don't know about calling native code from a thread? Should I be doing something explicitly to protect my native library's shared memory, signals, or semaphores?
Thanks,
Lisa.

Does the library do any initialization when loaded? Does it make a dangerous assumption that the thead calling cmdReadIt will always be the same thread that initially loaded it? Try loading the library in the worker thread instead of in the main thread via the static initializer.Yes. There is a call to another native method cmsOpenIt that does CMS initialization. I put cmsOpenIt as well as the System.loadLibrary("jastb") in the worker thread and I'm still experiencing the same problems.
Incidently, the mere presence of another Thread in this program does not seem to cause problems. It's only when I run calls to cmsReadIt in a thread that I get into trouble.
For example, this works fine:
public class CMSMessageHandler extends MessageHandler{
    public native int cmsReadIt(ByteBuffer buff);
    public ByteBuffer dirbuf = null;
    static {
        System.loadLibrary("jastb");
    public CMSMessageHandler() {
        // allocate to max size datagram
        dirbuf = ByteBuffer.allocateDirect(65536);
    public void readMessages(){
        try{
            int count = 0;
            while (true){
                count = cmsReadIt(dirbuf);
        } catch (Exception e){
    public static void main(String[] args) {
        CMSMessageHandler handler = new CMSMessageHandler();
        Thread worker = new Thread(){
            public void run(){
                while (true){
                    try{
                        Thread.currentThread().sleep(1000);
                        System.out.println("sleep thread.");
                    } catch (Exception e){
                        e.printStackTrace();
        worker.start();
        handler.readMessages();
    }Are there JVM flags that I should experiment with that might help mitigate/diagnose this problem?

How to 'STOP' a running java thread in J2ME?

Dear All,
How to 'STOP' a running java thread in J2ME?
In the middleware viewpoint, for some reasons we have to stopped/destroyed the running threads (we have no information how these applications designed).
But in J2ME, Thread.destroy() is not implemented. Are there other approaches to solve this problem?
Thanks in advance!
Jason

Hi jason,
Actually there are no methods like stop() and interrupt() to stop the threads in J2ME which is present in normally J2SE Environment.
But the interrupt method is introduced in Version 1.1 of the CLDC.
So, we can handle the thread in two ways.
a) If it is of single thread, then we can use a boolean variable in the run method to hadle it. so when the particular boolean value is changed , it will come out of the thread.
for eg:
public class exampleThread implements Runnable
public boolean exit = false;
public void run()
while(!exit)
#perform task(coding whatever u needed)
public void exit()
exit = true;
b) If it is of many threads then we can handle using the instance of the current thread using currentThread() method
for eg:
public class exampleThread implements Runnable
public Thread latest = null;
public Thread restart()
latest = new Thread(this);
latest.start();
public void run()
Thread thisThread = Thread.currentThread();
while( latest == thisThread )
#perform some tasks(coding part);
public voi d stopAll()
latest = null;
while ( latest == thisThread )
performOperation1();
if( latest != thisThread )
break;
performOperation2();
Regards,
Prathesh Santh.

Java Thread Referencing

What would be the best way to run a thread from a jsp and be able to check the status on that thread every once in a while?
I currently have a jsp page that instantiates a Stateful Session Bean and saves the handle of that SSB in the servlet context. From the SSB I initialize and start a thread. The thread does a batch process which ends up taking hours, so my jsp page automatically refreshes itself and pulls the handle to the SSB from the servlet context and calls methods that retrieve statistics on the thread. For example there is a method in the SSB called PercentDone() which in turn calls a method in the thread called PercentDone() which returns an integer.
My goal is to use the browser to start a batch process on the server and then monitor the progress of that batch process. I need the batch process to NOT be dependant at all on the browser, meaning I can shut it down and open it again and the process is still running and I can get a current stat on the process.
What I have now works, but I have heard that in later versions of Java it won't, for some reason. I am toying with the idea of using a deamon thread but don't know how to reference the thread after having started it.
I hope someone has some experience with something like this? I apreciate your time. Thank you

Create some sort of "Job" table, and when you create your long-running task (thread), associate it with an entry in the Job table. You will need to pass the job handle to the thread when (or before) you start() it, of course.
Your job thread should update the job status periodically, and your Job table should have methods to report the completion status and thread status of your thread. Don't have your thread unregister itself from the Job table when it's done, but instead, just set the thread handle in the Job table to null (otherwise you'll get a thread leak!), and periodically clean up old Job entries whose threads have died.
Such a solution is quite portable, and works quite well. And more importantly, even if you kill your browser and restart it, on the server side, since the job table is universal, any browser can go find the previously submitted jobs and report on their status.

Regarding Java Thread

Hi,
Iam having a problem......
My work is to process IP address one by one from a LOG file.
As the log file is very big so iam using Threads to reduce time.
Like if iam having 77 ip address then iam opening 10 connections
at a time by 10 thread...and the rest 7 ip are processed one by one..
But the thing is that when i start to run the application
the console window(JBILDER console window) goes off and next time it works or may the other time it works...i have to run again and again untill the out put comes............
if the log file is big then i have to run the application
2-3 times after than i get the console window stays..and gives the output..
But if iam not using the threads then the window nevers goes off
but the time to process is very high..3 hrs to process 100 ip address
jbuilder console window...
AppAccelerator(tm) 1.2.010 for Java (JDK 1.2), x86 version.
Copyright (c) 1997-1999 Inprise Corporation. All Rights Reserved.
robot.szukacz.pl
Company Name: 24 GODZINY SP. Z O.O.
street: UL. GARAZOWA 7
city: 02-651 WARSZAWA
location: PL
handle: nsk241211
phone: +48.226077884
Press Ctrl+C to terminate the application...
here as i run only one single ip so no problem ,
but i run a big log file 2 mb then this window goes off......
after trying 2-3 times then the window stays and gives the output..
thanks..
so can any one help me...
Ranjan...
U can mail me at [email protected]

Precisely what are you doing with a individual line which causes it to take so long ? If that isn't the result of a algorithm which is inefficient then you could possibly thread that portion out. You'd probably use some sort of thread pool, cause if you had 50000 ips then you'd be spawning 50000 threads. Something like 10-20 threads is ok I guess. So your log file parse gets a line and picks a available thread from the pool, if there is one free, and then send it off to the thread for processing. You need to gurantee that your thread can do it's job in polynomial time(that means in a reasonable amount of time) for this to be effective and consider how the main program is going to communicate with the threads.
I suggest first taking a look at your algorithm first before embarking on a threading quest (ie look for potential bottlenecks). It's just simpler (IMO) to handle.
ps. It's quite difficult to read responses in CAPS. Causes eyeaches. :(
>
YA YOU ARE RIGHT BUT IF WE GO BY SEQUENTIALLY THEN IT
WILL TAKE LONG TIME AS TAKE A IP ADDRESS FROM LOG FILE
AND FIND THE ADDRESS OF THE IP
LIKE 191.191.19.11 BELONGS TO SOME ONE .WE CAN FIND
OUT FROM ARIN.NET,
APNIC.NET OR RIPE.NET...
I HOPE U UNDERSTAND WHAT I MEAN TO SAY......
SO TO MAKE THINGS FAST IAM TRYING TO MAKE 10
CONNECTIONS AND
EACH THREAD HANDLES ONE CONNECTION.....
BUT WHEN I RUN FOR THE 1ST TIME THE JBUILDER CONSLOLE
WINDOW GOES OFF
AND AFTER TRYING AGAIN IT WORKS.....OR MAY AFTER ONE
MORE TIME.....
IF U STILL HAVE SOME DOUBTS THEN ASK ME
THANKS
RANJAN

Calling pl/sql api through multiple java threads

Hi All,
I need to call a pl/sql api from multiple java threads simultaneously and all thread will use same db connection.
I want to know if all the threads will simultaneously call the pl/sql api then will the local variable inside pl/sql procedure be shared between them or they will get separate instances of variables.
TIA

You cannot make multiple parallel client calls over the same Oracle session handle. There is a single non-threaded/non-fibre serialised server process servicing client requests for that session. (physical process on Linux/Unix, thread in the oracle.exe process on Windows).
Each thread on the client side, needs its very own Oracle session. Thus each thread will have a server process footprint on the server. Which could be problematic if 10 clients each starts 10 threads - as it means a 100 processes on the server are needed to service these client threads.
Have a look at Overview of OCI Multithreaded Development in the Oracle® Call Interface Programmer's Guide for how to use the threading call interface of the OCI - as oppose to rolling your own where each thread manually needs to deal with is OCI session context.

Creative Cloud Leaks Thread Handles

I have Creative Cloud version 1.8.0.447 running on windows 7, and I notice that it leaks a Thread Handle, about 4 times, or so, per minute. Since I leave my computer on most of the time, after a few days, the number of leaked Thread Handles is up in the tens of thousands.
I can see the thread handles leaking using Process Explorer.
Generally when a thread ends, its handle needs to be closed, or else the handle will stay active forever, or until the process ends. It looks like the programming team is not closing these handles after the threads end.
Please, adobe, fix this problem in a future release of creative cloud.
-Todd

Some 10.9.x links
-next link says After Effects, but check YOUR permissions !!!
-http://blogs.adobe.com/aftereffects/2014/06/permissions-mac-os-start-adobe-applications.ht ml
-Mac 10.9.4 and OpenCL issue https://forums.adobe.com/thread/1514717
-Mac 10.9.3 workaround https://forums.adobe.com/thread/1489922
-more Mac 10.9.3 https://forums.adobe.com/thread/1491469
-and https://forums.adobe.com/thread/1507936
or
A chat session where an agent may remotely look inside your computer may help
Creative Cloud chat support (all Creative Cloud customer service issues)
http://helpx.adobe.com/x-productkb/global/service-ccm.html

Java Threads not being released after loggin off

Hello everyone,
We are seeing a weird problem in our PI 7.0 box.
Once I logout from the XI box (both ABAP and Java stck) my basis tem still sees Java threads aginst my id still open. Our system does not seem to be realeasing Java threads.
Is this a know problem? What are the remidition steps.

Hi,
Are there existing HTTP sessions only?
The following might be helpful.
http://help.sap.com/saphelp_nw73/helpdata/en/c7/5ee440ba994fa3b187ff2f050cfe7c/content.htm
http://wiki.sdn.sap.com/wiki/display/ERPHCM/Sessionnotendingafterlogoff
Regards,
Varun

Java Threads Problem

Hi, I am trying to write a simple java threads program where in one thread reads a file and another thread writes the data into a second file....
Here is my code, although i think i am correct, my program still runs in a sequential fashion help help help!!!
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.FileOutputStream;
import java.io.IOException;
class MyThread extends Thread{
private int a;
private int c;
FileInputStream in;
FileOutputStream out;
public MyThread(int a){
this.a = a;
public void run(){
if(this.a==5)
     try {
     in = new FileInputStream("Britney.txt");
} catch (FileNotFoundException e) {
     // TODO Auto-generated catch block
     e.printStackTrace();
try {
     while((c=in.read())!=-1)
          a = (char) c;
          System.out.println(a);
catch (IOException e) {
     // TODO Auto-generated catch block
     e.printStackTrace();
if(this.a==10)
     try {
          out = new FileOutputStream("romi.txt");
     } catch (FileNotFoundException e) {
          // TODO Auto-generated catch block
          e.printStackTrace();
     for(int i = 0;i<50;i++)
          try {
               System.out.println(c);
               out.write(c);
          } catch (IOException e) {
               // TODO Auto-generated catch block
               e.printStackTrace();
class MainMyThread{
public static void main(String args[]){
MyThread thr1, thr2;
thr1 = new MyThread(5);
thr2 = new MyThread(10);
thr1.start();
thr2.start();
}

Encephalopathic wrote:
malcolmmc wrote:
... Chances of getting any kind of reply except "me too" can be pretty remote. ....there's actually a better chance of getting a reply on a general forum like this one, ....Can't you just post in both the narrow and the general forum, but include links one to the other in each thread? Or is that against the forum rules/etiquette? Most people here are ok with a crosspost IF those links are included.
I would also ask that the OP designate one of those threads as the real discussion thread and just direct folks there from the other threads, so that we have one coherent discussion. If he does that, there's no problem with trying to reach out to as broad an audience as possible.
I would figure that if the poster were upfront about what they are doing, folks wouldn't mind, but I could be wrong.I think that's generally the case. As long as the discussion is confined to one thread (and pointed there from the crossposts) or at the very least all the participants can see all the discussions, I think most people don't have a problem with it. It's when we waste our time answering when he's already got the answer elsewhere that's annoying.

Java thread not working in windows 7

I have a Java app that runs just fine on Windows XP. However, on Windows 7 there is a problem related to a job being done via a Java thread.
Background: The app spawns a separate job via a Java thread and continues on while the spawned job executes. In other words, 2 things happen at once.
Problem: On Windows 7, when the app spawns the job, Windows sits and waits for the spawned job to complete before it continues on. In other words, 2 things DO NOT happen at once.
It's as if Java threading does not work in Windows 7. Any input would be greatly appreciated. Thanks.

l_sleven wrote:
Thanks for the input, guys. From now on I'll be sure to include a SSCCE.
I'm beginning to believe this is actually my ignorance of threading in SWT, and XP was more forgiving than 7.
I would include a SSCCE, but of the 4 different ways I tried to get this to work I wouldn't know which to include since all failed to fix the problem.
If you want to identify a link to a web page identifying 'best practices' for SWT threading, that would be great. Be aware though that there's a good chance I've already googled the page.
A programmer smarter than me once said, "The road to best practices is a bumpy ride".[Concurrency in Swing|http://download.oracle.com/javase/tutorial/uiswing/concurrency/index.html]

Java thread with high CPU usage

Hi,
Running prstat -L -p pid on Solaris produces a list of lwp threads running in that process that are consuming the most CPU. Out of about 7 threads that were associated with kernel threads in pstack output, only 5 of them were mapped to a JVM thread dump. How is it possible to have a Java thread consuming CPU that doesn't show up in a thread dump?
Thanks,
J.S.

And, any idea what this thread is doing?
Here's pstack output for this missing thread 12:
----------------- lwp# 18 / thread# 12 --------------------
fe0e48b4 __1cHCompileSflatten_alias_type6kMpknHTypePtr__3_ (b6001780, 100360, fe4c8000, b6001780, 100360, 0) + 4d8
fe0f42ac __1cHCompilePget_alias_index6kMpknHTypePtr__I_ (b6001780, 100360, b6001264, 7e2bb8, 41f4, 7e2bb8) + 8
fe11131c __1cJStoreNodeFIdeal6MpnIPhaseGVN_pnLPhaseDefUse__pnENode__ (0, 7dd5b4, 5035ec, 7e2bb8, b6001144, b6001264) + 128
fe0c1224 __1cMPhaseIterGVNNtransform_old6MpnENode__2_ (0, b6001144, 7e2bb8, b6001264, 7e2bb8, b600116c) + 47c
fe17349c __1cMPhaseIterGVNIoptimize6M_v_ (24, 0, a37700, b6001110, b6001100, 3442f0) + b4
fe1943b0 __1cOPhaseIdealLoop2t6MrnMPhaseIterGVN_pk0_v_ (b6000ee8, a3772c, 1, 4ff37c, 2000, 269568) + 7cc
fe1cc89c __1cHCompileIOptimize6M_v_ (b6001780, b60015b8, b6001780, b60015dc, 0, b60013dc) + a0
fe1cb69c __1cHCompile2t6MpnFciEnv_pnHciScope_pnIciMethod_ill_v_ (91f504, b6001800, db8a24, fe5296b4, b60018a0, b60018b0) + 7bc
fe1c73f8 __1cKC2CompilerOcompile_method6MpnFciEnv_pnHciScope_pnIciMethod_il_v_ (27b18, b6001af8, db8a24, db8938, ffffffff, 1) + 70
fe1c79fc __1cNCompileBrokerZinvoke_compiler_on_method6FpnLCompileTask__v_ (db8938, db8a24, fe4eacec, 0, 0, 8d9) + 40c
fe280964 __1cNCompileBrokerUcompiler_thread_loop6F_v_ (28758, 13a570, fe4c8000, b6001d10, fe4c8000, ffffffff) + 168
fe216200 __1cKJavaThreadDrun6M_v_ (b5e02000, fe4d3e34, fe4c8000, 1ffd70, 13a570, 1ffd70) + 3d8
fe213ec8 _start (fe4c8000, ff325d10, 0, 5, 1, fe401000) + 20
ff36b734 threadstart (13a570, 0, 0, 0, 0, 0) + 40

I want to execute other java thread in the JNI.

Hi!
I would like to execute other java native thread concurrently in the Java Native Interface.
When a Java thread called a native thread to use java native interface, Java thread is blocking for processing native thread.
So, other java threads don't execute because they are blocking.
Why do these problems occur at JVM(personal basis profile:CVM)?
I hope you explain this reason in detail.
Regard.
Message was edited by:
peeppeep

Cross post
http://forum.java.sun.com/thread.jspa?threadID=786857&messageID=4471999#4471999

Best option for Asynchronous method invocation? JMS or Pure Java Thread

Hi,
We've a swing based Client application which is supposed to run with a server in J2EE environment. Some process like Search etc are very time consuming. So we are going for asynchronous process.
Now the question is to find a best option for this. Two possible candidates are
1. JMS
2. Java Thread.
Can anybody suggest me which one is the best option in this context?

Actually my thought was the issues with code maintainability.Maintainability is different issue, it is much related to OOAD and design pattern you might want to choose to avoid coupling (thus high reusability and maintainability - eg: for future enhancement, etc)..
public interface SearchService {
     public static class DefaultFactory {
             public SearchService getInstance() {
                      return HttpSearchService.getInstance();
     public void search (String[] keywords, Observer obs);
public class HttpSearchService implements SearchService {
     public static HttpSearchService getInstance() {
            ... bla bla bla singleton ...
     public void search(final String [] keywords, final Observer obs) {
                 (new Thread() {
                          public void run() {
                                .. do http request ...
                                List result = ... parse http response ...
                                obs.update(list);
                 }).start();
SearchService ss = SearchService.DefaultFactory.getInstance();
ss.search( new String[] { "get", "me", "my" , "dukes" }, myTableModel );
...You can, in future, replace the default factory HttpSearchService with something faster, more appropriate SearchService, without changing many codes.
rgds,
Alex

Java Thread Handling

Similar Messages

Maybe you are looking for