Strange NPE in the TransactionManager.forceResume()

Has anyone seen this exception?
          java.lang.NullPointerException
          at
          weblogic.transaction.internal.TransactionManagerImpl.getThreadProp(TransactionManagerImpl.java:1670)
          at
          weblogic.transaction.internal.TransactionManagerImpl.associateTxWithThread(TransactionManagerImpl.java:1640)
          at
          weblogic.transaction.internal.TransactionManagerImpl.internalResume(TransactionManagerImpl.java:1465)
          at
          weblogic.transaction.internal.TransactionManagerImpl.forceResume(TransactionManagerImpl.java:469)
          We get a reference on the Weblogic Transaction Manager in the server,
          call forceSuspend(), then do some DB work and when we try to resume the
          suspended transaction we get the NPE.
          This is in WLS 7.0 SP4 on Windows
          Thanks,
          Dejan

Your renderer is as I suspected the culprit.... I have made some quick changes (without testing):
class myCellRenderer extends DefaultTableCellRenderer {
   public Component getTableCellRendererComponent(JTable table, Object value, boolean isSelected, boolean hasFocus, int row, int column) {
      if (value == null ) setText("");
      setIcon( null );
      setBackground( Color.white );
      if (column != 0 ) {
         if (value instanceof Integer ) {
            int sevInt = ((Integer) value).intValue() ;
            setHorizontalAlignment( JCENTER );
            if (sevInt == 4 ) {
               setIcon( servDownImage );
               setText( "" );
            } else if (sevInt == 5 ) {
               setIcon( servUpImage );
               setText( "" );
      if ((row % 2) == 0) {
         setBackground(Color.white);
      } else {
         setBackground(new Color(229, 229, 229));
      if (isSelected) {
         setBackground(new Color(180, 180, 180));
      return this;
}Try it and see ....
;o)
V.V.

Similar Messages

Strange behaviour of the rangeSize - or is it desired?

Hi there,
as recently asked in Difference between VO Fetch Size, Iterator rangeSize and af:table Fetch Siz we have a problem with the rangeSize property of the iterator binding.
First of all, we're currently working wit JDev version 11.1.2.2.0. So, we have a page with a table. In this table is nested inside a panel collection which is nested inside a panel stretch layout. To cut a long story short, it stretches.
The rangeSize is now set to the default 25, but you could see in the fullscreen usage about 35 rows. When we now select a row which is in the rage of the first 25 records and use the data in such a method:
RichTable table = this.getTable();
RowKeySet rowKeySet = table.getSelectedRowKeys();
Iterator selectetNodesIterator = rowKeySet.iterator();
CollectionModel treeModel = (CollectionModel)table.getValue();
JUCtrlHierBinding treeBinding = (JUCtrlHierBinding)treeModel.getWrappedData();
while (selectetNodesIterator.hasNext()) {
List selectedRowKeyPath = (List)selectetNodesIterator.next();
JUCtrlHierNodeBinding nodeBinding = treeBinding.findNodeByKeyPath(selectedRowKeyPath);
Row row = nodeBinding.getRow();
DBSequence aaId = (DBSequence)row.getAttribute("AaId");
}We got a NPE at Row row = nodeBinding.getRow();, because the nodeBinding is null. Only the last fetched records have a JUCtrlHierNodeBinding.
So yeah ok, for testing purposes we changed the rangeSize to 1, just to see what happens. And again it fetches one row after another as long is the tables current view is full. And if we try it again every record results in a NPE excpet the very last visible record. So far so good, after what we've seen with 25 records this was expected.
When we change the rangeSize to a value greater than 36 (the visible rows, in that current fullscreen resolution) it works as it should.
So, to see if there is a bug i decided to create a small testapp with the HR schema. I've just dropped the employees vo as a table inside a panel collection which is nested inside a panle stretch layout. And used nearly the same method:
public String onButtonClicked() {
RowKeySet rks = table.getSelectedRowKeys();
CollectionModel tablemodel = (CollectionModel)table.getValue();
JUCtrlHierBinding binding = (JUCtrlHierBinding)tablemodel.getWrappedData();
Iterator it = rks.iterator();
while(it.hasNext()){
    List selectedRowKeyPath = (List)it.next();
    JUCtrlHierNodeBinding node = binding.findNodeByKeyPath(selectedRowKeyPath);
    Row row = node.getRow();
    System.out.println("Testausgabe: " + row.getAttribute(1));
return null;
}When the rangeSize is now set to 1 every visible row except the last also result in a NPE. If we run that above code on a button click, but now the strange part. When the rangeSize is changed to a value greater 1 it works like a charm. You can run that code with every visible record an you get now NPE. Also very strange with a rangeSize = 1, when you select the last visible record which is working and run the code, then select a record above, which had thrown an NPE previously and run the code. It is working, but with the value of the last visible record.
This leads us to a bunch of questions.
1) Is it the supposed way to work, that you get a NPE when the rangeSize is set to a smaller number than visible rows?
2) If 1) is the desired way, how do you handle the rangeSize for a table which is stretched? You couldn't now how many rows you could see.
3) Does anyone have an idea how to explain the last explained behaviour with the testapp? (I could upload it)
Thanks in advance.
Alex

Hi Alex,
can you open a SR at Oracle support for this issue?
I know there are several bugs which are somehow related with the rangeSize and the synchronization of the model (currrent row) and the view (in this case the visible part of af:table).
See also Restore current row after rollback does not work with page ranging
As far I know Oracle plans some enhancements in this area to fix the syncronize problems and change e.g. the rangeSize dynamically depending on the number of rows visible in the GUI.
But anyway your issue sounds like a new bug.
regards
Peter

How to get the TransactionManager in a stateful EJB (using JPA toplink)

I use weblogic 10.3
This is my stateful EJB and I need to get the TransactionManager because I want to suspend my transaction...
Here it is ok for the UserTransaction ut but I got an error with the TransactionManager
EJB Code
//@Stateless(name = "MyFacadeEJB", mappedName = "MyFacadeEJB")
@Stateful(name = "MyFacadeEJB", mappedName = "MyFacadeEJB")
@TransactionManagement(TransactionManagementType.BEAN)
@Remote
@Local
public class MyFacadeEJBBean implements MyFacadeEJB, MyFacadeEJBLocal, Serializable {
@Resource javax.transaction.UserTransaction ut;
@Resource javax.transaction.TransactionManager tm;
Error Message at the deploy EJB
Unable to deploy EJB: MyFacadeEJB from PocJTA-PEJB-ejb:
[EJB:011026]The EJB container failed while creating the java:/comp/env namespace for this EJB deployment.
weblogic.deployment.EnvironmentException: [EJB:010176]The resource-env-ref 'test.ejb.MyFacadeEJBBean/tm' declared in the ejb-jar.xml descriptor or annotation has no JNDI name mapped to it. The resource-ref must be mapped to a JNDI name using the resource-description element of the weblogic-ejb-jar.xml descriptor or corresponding annotation.
     at weblogic.ejb.container.deployer.EnvironmentBuilder.addResourceEnvReferences(EnvironmentBuilder.java:639)
     at weblogic.ejb.container.deployer.EJBDeployer.setupEnvironmentContext(EJBDeployer.java:247)

Chpruvos,
Hi, I ran into a similar issue when specifying WebLogic specific descriptors in ejb-jar.xml and the corresponding weblogic-ejb-jar.xml. I no longer use this approach in our example code - we just use annotations. see: http://wiki.eclipse.org/EclipseLink/Examples/JPA/WebLogic_Web_Tutorial#Tutorial_Source
My session bean is stateless - I don't know how much it will help with your statefull bean managed config but here are copies of the ejb.jar configuration xml I used to use. All the resource*ref elements (see the resource-description* element) ended up in the weblogic specific weblogic-ejb-jar.xml descriptor file.
ejb-jar.xml
<?xml version="1.0" encoding="UTF-8"?>
<ejb-jar xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://java.sun.com/xml/ns/javaee" xmlns:ejb="http://java.sun.com/xml/ns/javaee/ejb-jar_3_0.xsd" xsi:schemaLocation="http://java.sun.com/xml/ns/javaee [http://java.sun.com/xml/ns/javaee/ejb-jar_3_0.xsd]" version="3.0">
*<enterprise-beans>*
*<session>*
*<ejb-name>ApplicationService</ejb-name>*
*<ejb-class>org.eclipse.persistence.example.unified.business.ApplicationService</ejb-class>*
*<session-type>Stateless</session-type>*
++
++
++
++
*</session>*
*</enterprise-beans>*
</ejb-jar>
weblogic-ejb-jar.xml
<?xml version="1.0"?>
<weblogic-ejb-jar>
*<weblogic-enterprise-bean>*
*<ejb-name>ApplicationService</ejb-name>*
*<stateless-session-descriptor>*
*<pool>*
*<max-beans-in-free-pool>8</max-beans-in-free-pool>*
*<initial-beans-in-free-pool>2</initial-beans-in-free-pool>*
*</pool>*
*</stateless-session-descriptor>*
{color:#008000}<resource-description>
<res-ref-name>localOracle10RemoteDS</res-ref-name>
<jndi-name>localOracle10RemoteDS</jndi-name>
</resource-description>
<resource-env-description>
<resource-env-ref-name>localOracle10RemoteDS</resource-env-ref-name>
<jndi-name>localOracle10RemoteDS</jndi-name>
</resource-env-description>
{color}++
*</weblogic-enterprise-bean>*
</weblogic-ejb-jar>
thank you
/michael
[http://www.eclipselink.org|http://www.eclipselink.org/]
Edited by: michael_obrien on Feb 26, 2009 9:56 AM

Strange problem with the mail app on an iPad

I've got a bit of a strange problem with the mail app on an iPad - it's sending and receiving mail ok but won't let you delete mail. If you keep at it, it appears as if you have deleted mail (though takes a very long time as you just have to keep deleting it over and over again) - but the size of the app still keeps going up - reached 600mb so far, even though there is hardly any mail or attachments (goes up quite a lot each time you try to delete anything) - any idea what's causing this and how I can fix it?
It occasionally comes up with a dialogue panel saying to check my account settings, but I have checked those and they are fine. The little 'loading' icon just keeps spinning constantly too.

go to Settings > Mail > your account > Advanced > Move Discarded Messages Into
and make sure that Deleted Mailbox is ticked, and not Archive Mailbox
You can also adjust how long the deleted mail stays in the trash immediately below that.
If you are talking about a gmail account, you may also have to login to your account on the gmail server and update the settings there. This is what Barney previously posted about that:

Re: Satellite A200-1GB - strange ripples on the external LCD screen

H,
when I try to connect my laptop to my Samsung 32" LCD with VGA, I have strange ripples on the screen.
I tried changing every settings in my computer and in my TV and it's still there.
How can I fix this?
It's almost impossible to watch like this.
Thanks.

Did you test the notebook with other external monitors?
You should do this to ensure that this is not a LCD issue
Furthermore I recommend checking if such ripples are visible using different screen resolutions!
Last but not least an graphic card driver update would be useful too.

Strange "code" in the Comments field when playing songs

When I play some songs (probably 80% of my archive) iTunes generate some strange code in the Comments field. This is quite annoying, since I use the field to write additional info about the song, artist, location, whatever.
Examples of such codes:
" 00000244 00000000 000010B3 00000000 00022325 00000000 000057AD 00000000 00004E37 00000000"
" 0000014C 00000151 00000551 00000580 0001119E 0001119E 00003C73 0000379B 0001D4D7 0001D4D7"
and so on...
How can I stop this generation of random numbers and letters in the comments field?
null

Ick.... Well, maybe you're right. I also turned off "let itunes keep things organized" at about the same time. Between turning off sound check, itunes organizing, and the WMP settings, I have never seen it happen again.
I still think it was WMP, as I only open WMP rarely to listen to web clips and I noticed it had happened to some tracks I had not played in itunes for years, and it happened right after that WMP update. I'd had the sound check and itunes organization thing on for years.
One thing we can agree on, it was totally random. No rhyme or reason to it, and I even noticed that it didn't even affect all the tracks on some albums.
I just turned sound check back on, only on my Mac, because that is just my backup copy of files. We'll see if it hits again.
edit-(it's determining sound volume on x of 393, weird, I have over 27000 tracks in my library)
Message was edited by: Katrina S.

Many messages are showing up randomly on my mail.they say no message and no sender and give very strange dates. The text is gray - I can't delete them. What is this?

Many messages are showing up randomly on my mail. They say no message, no sender and have strange dates. The text is gray; I cannot delete them. What is this?

The Best Alternatives for Security Questions and Rescue Mail
    a. Send Apple an email request at: Apple - Support - iTunes Store - Contact Us.
    b. Call Apple Support in your country: Customer Service: Contacting Apple for
        support and service.
    c. Rescue email address and how to reset Apple ID security questions

I have recently bought a second I phone4 and have the same apple ID on both of them, strangely I noticed the messages sent from one phone appearing the other one along with any pictures I sent from one of my Iphones. Is it possible to see the same msgs.

I have recently bought a second I phone4 and have the same apple ID on both of them, strangely I noticed the messages sent from one phone appearing the other one along with any pictures I sent from one of my Iphones. Is it possible to see the same msgs.

You need to setup different Receive At addresses for iMessage.
If you don't want the same pics on each, turn off PhotoStream.
User Guide

I get a strange behavior of the tab bar and of the location bar in Firefox 29.0 for Mac.

I have just installed Firefox 29.0 for Mac.
I get a strange behavior of the tab bar and of the location bar with this new version.
Instead of the location bar, I get two rows of symbols. And it's impossible to write anything in the location bar.
(I'd like to add a screenshot, but I cannot find a way to do it.)

Thank you for your tip.
I found the culprit: it was an extension called RSS Icon 1.0.6.
I removed it and now Firefox 29.0 is working perfectly.
Now I'll have to find a replacement for that extension.
Thank you once again. Your tip was essential.

Strange behavior in the top link bar and open in a new tab

Hello,
I am having a strange behavior in the top link bar. Every element in the menu has the "open in a new tab"-option checked and is working fine but is not working at all in a sub site which inherit the menu.
I have look at the code thinking that the menu in the sub site did't get the target="_blank" but is there.
Anyone who knows what can be happening?
Thanks!!!!
Discusion actualizada

No worries, I found the problem.
Strange but Minimal download strategy feature cause the problem. I unactivated it and everything works fine.
Discusion actualizada

Strange character at the end of some file/folder names

I'm getting a strange character at the end of some of my file and folder names. This is it:  When I enlarge it, it says "private use" on the top and bottom, "E000" on the left side, and "F8FF" on the right side. It acts like a character within the font, but I'm not sure why it's there.
In case that didn't work, I'm going to try to attach or embed a gif of the character. I'm not sure if it will work or not.
Also, some longer file names seem to be dleeting some of the letters and are inserting a pound sign and a few numbers (e.g. #34) in place of the letters.
What's going on?
As usual, thanks in advance,
Lloyd

I thought that might be it...at first. Then I found some file names that were nowhere near 32 characters long whose names were truncated.
It's not a huge problem, except in QuarkXPress when I collect the files for output and I have to reconnect to the ones whose names have been changed (to protect the innocent, I suppose).
Thanks for the input mark.
Lloyd

Strange behavior of the notifications

Hi guys.
I use an iPhone 5S running iOS 7.0.3 and have a strange behavior of the notifications.
Example:
I get 3 messages from WhatsApp.
The lock-screen shows the preview of all 3 messages. OK so far.
But if I unlock my iPhone, a banner shows the latest message and the app-badge shows only 1.
This I have with other apps too.
This was not with older iOS versions
All settings looks OK. Do you have an idea?
Thank you, Tom

Because you are comparing floating point numbers.
For example 0.1 cannot be exactly represented in binary, so repetitve additions will probably not end in an exact integer later.
From the three inputs you can calculate the number of values exactly, and use a FOR loop. Don't use convoluted code....
Why don't you simply use ramp pattern?
LabVIEW Champion . Do more with less code and in less time .

Strange phenomena with the internal NIC

Hi!,
maybe there is someone with an idea or suggestion. I've got a PowerBook with a strange behavoir of the internal Onboard-NIC. Suddenly the connection to the internal Network stopped working.
But some things still work oO! If i plug in the Cable, the DHCP assigning works flawless, and i can ping all the interna, interaces without any Problems (AirPort, NIC, loopback). If i try to ping for example the router (from wich the PB gets the IP via DHCP) he says the Host is down.
Ok i've tried to manually change Link-Speed and switched between Half and Fullduplex mode, without any succes. The Book was resinstalled several times with Panther and Tiger and the connection via Airport works great. I cannot believe, that the NIC is completly damaged, because of great working DHCP... maybe some Problems in some OSI-Layers? Or because DHCP is not a layer3 proto?.... strange isn't it?
I've got the problem for a long time (a year or more) but i don't want to miss a 100Mbit conneciton because of Filetransfers to other PCs. Last year there wasn't one Ehternet PC-Card available that works with panther oder tiger Can someone give me a suggestion for a card that works with 10.3 or greater and is available in germany?
thanks a lot
--edit ---
find another strange behavoir... if i disable all other interfaces, and trying to get the arp table for en0, there is no listing! The ARP "Hangs", can only CTRL-C to come back to console... "sudo arp -a".. (if i wait very long, the entries will show up (about 30sec for each entry)....
When do the same with only Airport enabled, sudo arp -a gives me all the information i need!
strange?
I've added a permanent ARP-Entry for the router.. wihtout success...

Well, I can't offer any suggestions, but I also see the same behavior on my PowerBook. I have no idea what has changed, or where to find clues, but earlier this evening I came back to my computer (already running and awake) to find this odd behavior. I spent more than an hour trying different things, and all I could get was a new ip from the dhcp server (although I couldn't ping it from Terminal window!).
The behavior for Airport and Ethernet was the same - obtain an ip address from the dhcp server, but no network traffic was possible. And yes, I verified that the network was working fine (other computers were online no problems).
After much fiddling around, I restarted my PoweBook, and whaddya know, everything is normal again.
Are there others who are experiencing this problem?
15" PowerBook G4 1.67 and G5 1.8 Mac OS X (10.4.7)

Strange output of the PID module

Dear all,
I meet a problem when using the PID module. I am only set P and use the PID as a linear gain. I found in some of my VIs , the output of the PID module is the opposite value, while in other VIs, the output is normal. In these case, I used to set the linearity of the PID as -1. Even though the VI can run normally in this way, I want to figure out what shall be the wrong about the PID module. Could you help me explain about the strange behavior about the PID module?
Please see the attached file, I add some probes like 24 and 25 to show the values that represent this problem.
Thanks
best
Attachments:
PID module opposite output.png ‏193 KB

Your set point is set to zero, your process variable appears to be 8.181 i.e. too high and error (SP-PV) is negative, so the output of your controller should be negative (-8.181 x Kp). The controller output is clamped to be between 0.5 and -0.5, and it is -0.5V, which is what you would expect if your Kp is the 0.1 one (I can't see which set of gains it is).
So it appears to be correct ?
Note: linearity should be between 0 and 1 (read the manual). Keep this set to 1 until you know things are working and only change if you need nonlinearity with error.
Consultant Control Engineer
www-isc-ltd.com

The strange behaviour of the notification

Coherence 3.3.1/389
     .Net API 3.3.1.2
     I have C++ GUI client with grid. It shows in the grid all updates from Coherence.
     I have also the 'server' application, which inserts objects into Coherence (1 - 2 per second).
     I use 'near-scheme' with front set to local-scheme and back set to remote-cache-scheme.
     I have 3 GUI clients connected to the Coherence and i have the strange behaviour of the clients. The insert comes to one of them three times instead of one. I've checked in the debugger and it's true. The insert really comes three times. But when i've restarted the client all became to work fine.
     It's bad that i can't repeat this situation, it happens from time to time.
     Does anybody get the same situation?
     I've masked it by the checking key and version of object, but it's really strangely.
     Thanks.

I think it's depended on the situation described in thread CacheEventFilter: Where does the filter part apply?
     Many thanks.

Strange NPE in the TransactionManager.forceResume()

Similar Messages

Maybe you are looking for