Multiple blocking locks on same object

we have a database that is providing a great deal of application 'freezing' due to lock problems that were not experienced before moving to a new 10g database environment.
one of the items i believe to be of interest is the fact that when this 'freezing' occurs then a look at the database locks reveals that more than one TX exclusive blocking lock is held on the same database object without any waits showing.
my understanding is that this should not be the case? that only one user should possess the exclusive transaction level lock and that the other user should only have an outstanding request for this same level lock?
any information you can provide would be most appreciated.
thanks in advance.

You would need to find out the sql the application is executing when those locks freeze like that. From there determine if there is a logical syntax issue or get the explain plan to ensure it isn't for example an update statement on a massive table that has improper indexing.
Also, do you see other sessions blocking other sessions as orafad suggested? In this scenario my original explanation is different. If the locks are all chained on waiting on another session, that means the first session that is blocking is updating the same data that the waiters are wanting.
In my opinion you need to find out the exact SQL causing these locks to ensure it is an issue with one of your applications. Determining the performance of that query is a must.
The only weird thing is you say there is no waits taking place....did you look in v$session_wait?
Lee

Similar Messages

Multiple Transports for the Same Object.

I have ABAP program 123 and I make a change, create transport A and release the transport. I do not request transport A go to production.
I then modify ABAP program 123 again, create transport B and release the transport. I do not request transport B go to production.
Now I request transport A go to production.
Can someone explain to me why the changes in transport B are included when transport A is moved to production?
I thought each transport would be at its own change level.
Thank-you.

I can guarantee all of you that when Transport A was moved, it contained changes that were made in Transport B.
Surely you understand that Thomas is correct and that there's a logical reason, not related to transport layer configuration, for the issue? You CAN have multiple transports open for the same object without a lock by forcing it but check the dates as already suggested...

Can we have multiple transports for the same object.

Hi guys,
Can we have multiple transports for same object in dev system. Can anyone tell me how can this be done.
Thanks

Its not possible for the same development object. Only 1 person can access an object at a time and if mutiple users modify an object new TASKs are created under the same TRANSPORT.
Only after releasing the tr you can create a new tr on the same object.
Message was edited by:
Abhishek Jolly

Multiple inserts for the same object

We have observed that in certain cases, the same object is attempted to be inserted twice, resulting in Primary Key violation.
Here is one example -
A(1->1)B, unidirectional, A has the foreign key to B
new B
copyB = register(B) (also assign sequence number)
new A1, A2
A1.set(copyB), A2.set(copyB)
register(A1), register(A2)(also assign sequence number)
commit
it tries to insert the same B twice.
any clues whats wrong here?
thanks

Any chance your 1:1 from A to B is marked as privately-owned?
This would indicate to TopLink that the object in each relationship is unique and should be inserted.
Doug

Multiple static synchronized methods locking the same object ?

If I have multiple static synchronized methods in a class will all the methods lock on the same (Class) object ? I guess the answer to this would be yes. In that case is it possible achieve synchronization without an object ie code level synchronization ? If yes, how ?

If I have multiple static synchronized methods in a
class will all the methods lock on the same (Class)
object ? I guess the answer to this would be yes. In
that case is it possible achieve synchronization
without an object ie code level synchronization ? If
yes, how ?There is nothing special about static synchronized methods vs. any other synchronization. It's just shorthand for syncing on the Class object for that class. The effects of synchronization are identical no matter what you sync on.
There's no way to achieve synchronization without an object. Synchronizing always obtains an object's lock. In the case of static synced methods, the object whose lock we obtain is the Class object for that class. As far as syncing goes, it's just another object.
There is no such thing as "code level synchronization" vs. any other kind. All syncing occurs in your Java code over blocks of code. For synced methods, those blocks are the entire respective method bodies.
Having said all that, I have no idea what you're really after or what you're trying to achieve.

I'm having problems in a GOOP application with multiple instances of the same object running simultaneously.

I?ve written an application to control a system with multiple lamps. Each lamp in the system has its own power supply. Each lamp object contains a reference to its power supply. The application contains an array with the references to all the lamps in the system. Each lamp is in charge of monitoring itself and setting its power supply?s voltage according to the applications commands, therefore each lamp needs that its ?run me? method will run as long as the application is running i.e. the run method has to be reentrant. Because there are multiple lamps and because the ?run me? methods for each lamp run simultaneously, to over come this problem each lamp has a met
hod ?run? that calls ?run me? with ?RUN VI? invoke node with the ?Wait until done? set to false. When I run the application, it seems to work but if I check the AO there is no change in the output, and the lamps seem to be reading their input from the same channel. I can?t seem to find the problem, I think that the fact that the method is reentrant causes some sort of mix-up in the separate object references. Also because the method is reentrant I can?t debug it (with one lamp it seems to work ok). Does anyone have any idea what could be causing this problem.
P.S I'm having problems attaching my sample program to this message. You can contact me at [email protected] to recive the sample program I wrote.

Jean, Thanks for your answer, I've tried checking the invoke node that calls the "Run VI" method but it doesnt return an error any of the times its called even that the Vi that it called is still running. I also tried your solution of changing the object's "Run" method to a template but that doesnt work either. I can't seem to post the sample progam that demonstrates the problem, is it possible to send it to you by email, contact me at [email protected]
Attachments:
Sample_program.zip ‏926 KB

Returning multiple instances of the same object?

I have a fairly straight forward swing test class that contains a combobox. The problem is that an object is not selected in the combobox because I get different objects when building the list and when setting the selected item.
To test this, I've added the following set of 4 lines on three locations in the code:
- immediately after Toplink is initialized
- when setting the selected item
- when the main exits (and the JFrame is open)
Stand s = Stand.findByPK(4).getDerrivedfromstand();
System.out.println( s + ", Using EM #" + Integer.toHexString( lEntityManager.hashCode() ) );
s = Stand.findByPK(1);
System.out.println( s + ", Using EM #" + Integer.toHexString( lEntityManager.hashCode() ) );
I happen to know that Stand 4 is derrived from Stand 1. So I get Stand 1 once via Stand 4 and once directly. The result is:
java.vm.version=1.6.0-b105
[TopLink Info]: 2007.05.30 11:01:10.531--ServerSession(33414193)--TopLink, version: Oracle TopLink Essentials - 2.0 (Build b41-beta2 (03/30/2007))
[TopLink Info]: 2007.05.30 11:01:11.343--ServerSession(33414193)--file:/C:/Documents%20and%20Settings/toeu/My%20Documents/kp/reinders/reinders/voorraad/container/bm/build/-reinders login successful
// immediatly after Toplink init
[1] setObject=4 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
[1] setObject=1 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
nl.reinders.bm.Stand@1866417, Standnr=1, Using EM #1a8e53c
[1] setObject=1 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
nl.reinders.bm.Stand@1866417, Standnr=1, Using EM #1a8e53c
// when filling a combobox in the GUI
executeQuery: SELECT ... FROM stand ORDER BY standid ASC
getElementAt 0 = nl.reinders.bm.Stand@1866417, Standnr=1
// when setting the value of the combobox
executeQuery: SELECT ... FROM stand ORDER BY standid ASC
[1] setObject=4 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
[1] setObject=1 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
nl.reinders.bm.Stand@171ef98, Standnr=1, Using EM #1a8e53c
[1] setObject=1 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
nl.reinders.bm.Stand@171ef98, Standnr=1, Using EM #1a8e53c
// when exiting the main (JFrame is open)
[1] setObject=1 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
nl.reinders.bm.Stand@171ef98, Standnr=1, Using EM #1a8e53c
[1] setObject=4 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
nl.reinders.bm.Stand@171ef98, Standnr=1, Using EM #1a8e53c
[1] setObject=1 / SELECT ... FROM stand WHERE (standnr = >>>HERE<<< )
executeQuery: SELECT ... FROM stand WHERE (standnr = ?)
nl.reinders.bm.Stand@171ef98, Standnr=1, Using EM #1a8e53c
What is amazing is that somewhere after the combobox is filled, the query returns another instance of the same record and thus the select item will not find an "equal" object in the list. Any suggestions?

Nevermind. A clear was done on the EM somewhere down the line.

Multiple instantiations of the same object

Can someone please explain to me the difference of these two circumstances upon the JVM:
1) I declare an object outside the loop and instantiate a new object on each iteration of the loop
Object obj = null;
for (int i = 0; i < 6; i++) {
    obj = new Object();
}2) I declare and instantiate the object inside the loop only
for (int i = 0; i < 6; i++) {
    Object obj = new Object();
}I assume in the first example that I am reserving memory space and overwriting it on each iteration, whereas in the second example I am reserving and consuming a new memory space on each iteration.
If that assumption is correct, which is the more efficient? Wouldn't option 1 be easier on garbage collection?

I compiled both cases using Sun javac (1.4.1) and used javap to view the compiled code. It looks the same to me. Here's the compiled code (for the loop) for the first case   2 iconst_0
   3 istore_2
   4 goto 18
   7 new #2 <Class java.lang.Object>
10 dup
11 invokespecial #1 <Method java.lang.Object()>
14 astore_1
15 iinc 2 1
18 iload_2
19 bipush 6
21 if_icmplt 7
24 returnHere's the compiled code the second case   0 iconst_0
   1 istore_1
   2 goto 16
   5 new #2 <Class java.lang.Object>
   8 dup
   9 invokespecial #1 <Method java.lang.Object()>
12 astore_2
13 iinc 1 1
16 iload_1
17 bipush 6
19 if_icmplt 5
22 return

Finalize() method being called multiple times for same object?

I got a dilly of a pickle here.
Looks like according to the Tomcat output log file that the finalize method of class User is being called MANY more times than is being constructed.
Here is the User class:
package com.db.multi;
import java.io.*;
import com.db.ui.*;
import java.util.*;
* @author DBriscoe
public class User implements Serializable {
    private String userName = null;
    private int score = 0;
    private SocketImage img = null;
    private boolean gflag = false;
    private Calendar timeStamp = Calendar.getInstance();
    private static int counter = 0;
    /** Creates a new instance of User */
    public User() { counter++;
    public User(String userName) {
        this.userName = userName;
        counter++;
    public void setGflag(boolean gflag) {
        this.gflag = gflag;
    public boolean getGflag() {
        return gflag;
    public void setScore(int score) {
        this.score = score;
    public int getScore() {
        return score;
    public void setUserName(String userName) {
        this.userName = userName;
    public String getUserName() {
        return userName;
    public void setImage(SocketImage img) {
        this.img = img;
    public SocketImage getImage() {
        return img;
    public void setTimeStamp(Calendar c) {
        this.timeStamp = c;
    public Calendar getTimeStamp() {
        return this.timeStamp;
    public boolean equals(Object obj) {
        try {
            if (obj instanceof User) {
                User comp = (User)obj;
                return comp.getUserName().equals(userName);
            } else {
                return false;
        } catch (NullPointerException npe) {
            return false;
    public void finalize() {
        if (userName != null && !userName.startsWith("OUTOFDATE"))
            System.out.println("User " + userName + " destroyed. " + counter);
    }As you can see...
Every time a User object is created, a static counter variable is incremented and then when an object is destroyed it appends the current value of that static member to the Tomcat log file (via System.out.println being executed on server side).
Below is the log file from an example run in my webapp.
Dustin
User Queue Empty, Adding User: com.db.multi.User@1a5af9f
User Dustin destroyed. 0
User Dustin destroyed. 0
User Dustin destroyed. 0
User Dustin destroyed. 0
User Dustin destroyed. 0
USER QUEUE: false
INSIDE METHOD: false
AFTER METHOD: false
User Dustin destroyed. 1
User Dustin destroyed. 1
User Dustin destroyed. 1
User Dustin destroyed. 1
USER QUEUE: false
INSIDE METHOD: false
AFTER METHOD: false
User Dustin destroyed. 2
User Dustin destroyed. 2
User Dustin destroyed. 2
User Dustin destroyed. 2
User Dustin destroyed. 2
User Dustin destroyed. 2
User Dustin destroyed. 2
User Dustin destroyed. 2
USER QUEUE: false
INSIDE METHOD: false
AFTER METHOD: false
User Dustin destroyed. 3
User Dustin destroyed. 3
User Dustin destroyed. 3
User Dustin destroyed. 3
User Dustin destroyed. 3
User Dustin destroyed. 3
User Dustin destroyed. 3
User Dustin destroyed. 3
User Dustin destroyed. 3
USER QUEUE: false
INSIDE METHOD: false
AFTER METHOD: false
User Dustin destroyed. 4
User Dustin destroyed. 4
User Dustin destroyed. 4
User Dustin destroyed. 4
User Dustin destroyed. 4
User Dustin destroyed. 4
User Dustin destroyed. 4
User Dustin destroyed. 4
User Dustin destroyed. 4
USER QUEUE: false
INSIDE METHOD: false
AFTER METHOD: false
User Dustin destroyed. 5
User Dustin destroyed. 5
User Dustin destroyed. 5
User Dustin destroyed. 5
User Dustin destroyed. 5
User Dustin destroyed. 5
User Dustin destroyed. 5
User Dustin destroyed. 5
User Dustin destroyed. 5
USER QUEUE: false
INSIDE METHOD: false
AFTER METHOD: false
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
User Dustin destroyed. 6
Joe
USER QUEUE: false
INSIDE METHOD: false
AFTER METHOD: false
User Dustin pulled from Queue, Game created: Joe
User Already Placed: Dustin with Joe
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
User Dustin destroyed. 7
INSIDE METHOD: false
INSIDE METHOD: false
USER QUEUE: true
INSIDE METHOD: false
INSIDE METHOD: false
User Dustin destroyed. 9
User Joe destroyed. 9
User Dustin destroyed. 9
User Dustin destroyed. 9
User Dustin destroyed. 9
User Dustin destroyed. 9
INSIDE METHOD: true
INSIDE METHOD: false
USER QUEUE: true
INSIDE METHOD: false
INSIDE METHOD: false
INSIDE METHOD: true
INSIDE METHOD: false
USER QUEUE: true
INSIDE METHOD: false
INSIDE METHOD: false
It really does seem to me like finalize is being called multiple times for the same object.
That number should incremement for every instantiated User, and finalize can only be called once for each User object.
I thought this was impossible?
Any help is appreciated!

Thanks...
I am already thinking of ideas to limit the number of threads.
Unfortunately there are two threads of execution in the servlet handler, one handles requests and the other parses the collection of User objects to check for out of date timestamps, and then eliminates them if they are out of date.
The collection parsing thread is currently a javax.swing.Timer thread (Bad design I know...) so I believe that I can routinely check for timestamps in another way and fix that problem.
Just found out too that Tomcat was throwing me a ConcurrentModificationException as well, which may help explain the slew of mysterious behavior from my servlet!
The Timer thread has to go. I got to think of a better way to routinely weed out User objects from the collection.
Or perhaps, maybe I can attempt to make it thread safe???
Eg. make my User collection volatile?
Any opinions on the best approach are well appreciated.

How can multiple threads hold the same lock?

I was always under the impression that no 2 threads could hold a lock for the same object, and yet the HotSpot traces say this is happening all the time.
For example Object 0x0a6c1bf8 is locked in 2 threads:
"10.129.24.35-1" daemon prio=6 tid=0x2836bad0 nid=0x9d4 runnable [0x2b28f000..0x2b28fc9c]
at java.net.PlainSocketImpl.socketConnect(Native Method)
at java.net.PlainSocketImpl.doConnect(PlainSocketImpl.java:333)
- locked <0x2493f6e0> (a java.net.SocksSocketImpl)
at java.net.PlainSocketImpl.connectToAddress(PlainSocketImpl.java:195)
at java.net.PlainSocketImpl.connect(PlainSocketImpl.java:182)
at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:366)
at java.net.Socket.connect(Socket.java:519)
at org.jacorb.orb.factory.PortRangeSocketFactory.doCreateSocket(PortRangeSocketFactory.java:142)
at org.jacorb.orb.factory.AbstractSocketFactory.createSocket(AbstractSocketFactory.java:54)
at org.jacorb.orb.iiop.ClientIIOPConnection.createSocket(ClientIIOPConnection.java:313)
at org.jacorb.orb.iiop.ClientIIOPConnection.connect(ClientIIOPConnection.java:160)
- locked <0x0a6c1b98> (a org.jacorb.orb.iiop.ClientIIOPConnection)
at org.jacorb.orb.giop.GIOPConnection.sendMessage(GIOPConnection.java:934)
- locked <0x0a6c1bf8> (a java.lang.Object)
at org.jacorb.orb.giop.GIOPConnection.sendRequest(GIOPConnection.java:900)
at org.jacorb.orb.giop.ClientConnection.sendRequest(ClientConnection.java:323)
at org.jacorb.orb.giop.ClientConnection.sendRequest(ClientConnection.java:304)
at org.jacorb.orb.Delegate.invoke_internal(Delegate.java:1024)
- locked <0x0d434a88> (a java.lang.Object)
at org.jacorb.orb.Delegate.invoke(Delegate.java:939)
at org.omg.CORBA.portable.ObjectImpl._invoke(ObjectImpl.java:80)
at com.company.MyApp._StateChangeListenerStub.newChanges(_StateChangeListenerStub.java:33)
at com.company.MyAppManager.ListenerOneway.send(ListenerOneway.java:65)
at com.company.MyAppManager.Sender.run(Sender.java:139)
"ClientMessageReceptor840" daemon prio=6 tid=0x2a3ab988 nid=0x6b4 in Object.wait() [0x2f70f000..0x2f70fb1c]
at java.lang.Object.wait(Native Method)
at java.lang.Object.wait(Object.java:474)
at org.jacorb.orb.giop.GIOPConnection.waitUntilConnected(GIOPConnection.java:278)
- locked <0x0a6c1bf8> (a java.lang.Object)
at org.jacorb.orb.giop.GIOPConnection.getMessage(GIOPConnection.java:318)
at org.jacorb.orb.giop.GIOPConnection.receiveMessages(GIOPConnection.java:464)
at org.jacorb.orb.giop.MessageReceptor.doWork(MessageReceptor.java:71)
at org.jacorb.util.threadpool.ConsumerTie.run(ConsumerTie.java:61)
at java.lang.Thread.run(Thread.java:595)
Have I miss-interpreted the HotSpot stack trace?

To summarise they are both in a synchronized block on the same mutex/Object, but one is in a wait and one isn't.
So then it follows that if the HotSpot thread reads " - waiting on <address>" like:
"SomeThread" daemon prio=6 tid=0x29a06658 nid=0xf28 in Object.wait() [0x2b18f000..0x2b18fa1c]
at java.lang.Object.wait(Native Method)
- waiting on <address> (a javax.swing.TimerQueue)
at com.comp.SomeThread.run(TimerQueue.java:236)
- locked <0x09539510> (a javax.swing.TimerQueue)
at java.lang.Thread.run(Thread.java:595)
then it means that the thread is trying to enter the synchronized block on the same mutex/Object but it can't because there is another thread in there which is running, so it is waiting on the outside of the lock instead of the inside of the lock like "ClientMessageReceptor840".
Ok, I think I get it now. Thanks everyone for your help.

How do I use multiple classes to access the same object?

This should be staightforward. I have and object and I have multiple GUIs which operate on the same object. Do all the classes creating the GUIs need to be inner classes of the data class? I hope this question makes sense.
Thanks,
Mike

public final class SingletonClass {
private static SingletonClass instance;
private int lastIndex = 10;
public final static SingletonClass getInstance()
if (instance == null) instance = new SingletoClass();
return instance;
public int getLastIndex() {
return lastIndex;
}1) This won't work since one could create a new SingletonClass. You need to add a private constructor. Also, because the constructor is private the class doesn't have to be final.
2) Your design isn't thread-safe. You need to synchronize the access to the shared variable instance.
public class SingletonClass {
    private static SingletonClass instance;
    private static Object mutex = new Object( );
    private int lastIndex = 10;
    private SingletonClass() { }
    public static SingletonClass getInstance() {
        synchronized (mutex) {
            if (instance == null) {
                instance = new SingletonClass();
        return instance;
    public int getLastIndex() {
        return lastIndex;
}if you are going to create an instance of SingletonClass anyway then I suggest you do it when the class is loaded. This way you don't need synchronization.
public class SingletonClass {
    private static SingletonClass instance=new SingletonClass();
    private int lastIndex = 10;
    private SingletonClass() { }
    public static SingletonClass getInstance() {
        return instance;
    public int getLastIndex() {
        return lastIndex;
}If you don't really need an object, then you could just make getLastIndex() and lastIndex static and not use the singleton pattern at all.
public class SingletonClass {
    private static int lastIndex = 10;
    private SingletonClass() { }
    public static int getLastIndex() {
        return lastIndex;
}- Marcus

I have multiple devices with the same cerificate, once I have a an exception for one, FF denys access to the others. I cannot change the certificates, I need to stop FF blocking them.

I have multiple devices with the same cerificate, once I have an exception for one, FF denys access to the others. I cannot change the certificates, I need to stop FF blocking them.
== This happened ==
Every time Firefox opened
== I attempt to access a web front end on an IBM SVC device

You can't use the same certificate more than once.
See also [[Certificate contains the same serial number as another certificate]]

Multiple data blocks on the same canvas

Forms newbie question:
Is it possible to have 2 data blocks with two different sets of transactional triggers (ON-UPDATE, ON-LOCK, etc.) on the same canvas?
I've got an example where i've got two data blocks (one sourced from a view, and one that is a CTL block) on the same canvas. The block sourced from the view is working fine. When I make updates, they are reflected, etc. But, i've got one field in the control block that I need to update from as well, and I can't seem to get it (or any of the CTL block items) to show up as "updateable", regardless of the set_item_property....theyre always grayed out.
Do i need to take the user to a new canvas to be able to utilize the update of the ctl field?

A second canvas is not needed. Normally, control-block items are always updateable, so there is something going on in your form to prevent it.
> I can't seem to get it (or any of the CTL block items) to show up as "updateable",
regardless of the set_item_property....theyre always grayed out.
What... Is there code someplace in the form that sets them to Enabled, false? If that is the case, then to get them working again, besides setting Enabled to true, you also must set Updateable and Navigable to Property_True. (This is documented at the end of the on-line help on Set_Item_Property.)

URGENT: Insert multiple rows in a view object cache from the same Page

Please let me know whether I can get insert multiple rows in the same view object from a single page.
I'm using UIX on Jdeveloper 9.0.5.2
Thanks a lot.

Hi,
please check Duncan Mills how-to on this
http://www.oracle.com/technology/products/jdev/tips/mills/JSP_Multi_Row_Edits.html
Frank

Object or Statement Block locking using synchronized keyword in java

Hi ,
Can anyone tell me how java is implementing the Object locking and the code block locking mechanism using the 'synchronized' keyword internally??
Thanks in advance
Raj

The JVM implements a concept known as a monitor using a combination of direct implementation and support by the operating system. For more detail, refer to the Java JVM and language specifications.
Chuck

Multiple blocking locks on same object

Similar Messages

Maybe you are looking for