|
Change Request
Number
|
Description
|
|
CR231621
|
DB2 XA Configurations ran at a slow rate while using DB2 XA configurations. There is a negative impact on performance if the application has been running for an extended period of time.
The problem has been resolved.
|
|
CR238789
|
Problem: Every call to setTaskCompletionDueDate on a task resulted in a JMS message posted to the wli.internal.worklist.timer.queue. This message is delivered at a later date and triggers callback calls on Task controls.
The Task controls list that waits for callbacks on a Task is in table WLI_WORKLIST_LISTENER. In the absence of a listener waiting on a task, every time setTaskCompletionDueDate is called, one more message is posted on this queue resulting in millions of tasks opened at a given time.
The problem has been resolved. WebLogic Integration posts expiration messages only if there are listeners waiting on tasks
|
|
CR252921
|
Problem: In some cases the XML to MFL transformation failed with a DataMismatchException as the output size of the MFL object did not match the required data definition.
The problem has been resolved. NonXMLWriterVisitor uses default values when these values are not specified.
|
|
CR247924
|
Problem: The Quality of Service of logging of tracking events to Runtime/Reporting data streams was not configurable. racking & Reporting can be configured as orUnrel
The problem has been resolved. If both data streams are configured as reliable, then the runtime tracking data and reporting data is always consistent with JPD. If any exception occurs during the flushing of these data, the entire transaction is rolled back.
If both data streams are configured as unreliable then the runtime tracking and reporting data is flushed in a different transaction. This data is independent of a JPD transaction. So in this case if the server crashes after completion of JPD, all the tracking and reporting data is lost.
|
|
CR256518
|
Problem: The WebLogic Integration cluster did not work if the Administration server is down.
The problem has been resolved.
|
|
CR258928
|
Problem: The current JMS connections continue to increase with a gradual increase in consumption of memory - and the current JMS connections are not properly closed due to faulty process tracking.
The problem has been resolved. Process tracking does not cause a JMS connection leak.
|
|
CR258935
|
Problem: The issues are resolved using the same method as default delimiters are resolved in Non-XML to XML transformation
|
|
CR260348
|
Problem: The pause time increases exponentially as more controls are added to JPD.
The problem has been resolved. WebLogic Integration caches icons to improve performance.
|
|
CR260893
|
Problem: When a JPD application was redeployed the user submitted a taskComplete action. The onTaskCompleted call back in the JPD instance did not receive a call and remained in a running state.
The problem has been resolved. RemoteWorklistManager throws an exception and marks the transaction for rollback if an error is encountered when despatching events for task callbacks.
|
|
CR262025
|
The issue of time calculation with respect to time zones has been resolved.
|
|
CR262179
|
Problem: At times, a stateful JPD does not persist.
This issues is now resolved
|
|
CR263031
|
Problem: The "PROCESS_INSTANCE" column of the WLI_TASK_ARCHIVING table logs the URI of the process, rather than the process instance ID.
The problem has been resolved. WLI_TASK_ARCHIVING.PROCESS_INSTANCE value matches corresponding value from WLI_WORKLIST_TASK.PROCESS_INSTANCE for stateful JPD.
|
|
CR265208
|
Problem: An ArrayIndexOutOfBounds exception is observed during the processing of an incoming MFL to the WLI process resulting in a large CSV file.
The problem has been resolved.
|
|
CR266540
|
Problem: When an XML document with a DAT element as root is submitted, it fails the runtime validation. The W3 documentation does not indicate that the included elements are considered local elements.
The problem has been resolved. WLI provides custom mapping for the prologue.
|
|
CR266580
|
Problem: When a new task was created in the callback path, TaskControl set the owner of the task to anonymous even though JPD was started by the WebLogic user. This occurs when the customer uses the task control factory. The task owner is the user who starts the JPD.
In both cases context.getCallerPrincipal() method returns "anonymous" opposed to user "weblogic"
The problem has been resolved.
|
|
CR269507
|
Problem: Transformation failed when a MFL document had RepeatUnboundedStructNode ending with a shared delimiter
When there was a RepeatUnbounedStructNode that was terminated by a shared delimiter of the parent Struct node, the transformation failed with DataMismatchException.
The problem has been resolved. NonXMLReaderVisitor and NonXmlWriterVisitor have been updated.
|
|
CR269796
|
Problem: When an ebXML control was wrapped within a Java custom control, on invocation from within a JPD, EBXMLControlHelper threw a ClassCastException.
The problem is resolved if you iterate through parent containers until the JPDContainer is found
|
|
CR272338
|
Problem: An IndexOutOfBoundsException was thrown from internal source code. onException received an IndexOutOfBoundsException from an internal source code. This occurs right after a sub process finished its work then called back a parent process.
The problem has been resolved.
|
|
CR274066
|
Problem: When a dynamic transformation control was used, and a URL (referencing a file containing MFL) in the code to perform transformations, if the MFL file was updated, that update was not effective unless the server was restarted.
The problem has been resolved. The unregisterMFL method has been added to dynamic transformation control
|
|
CR275105
|
Problem: A NumberFormatException occurred during a Bin to XML transformation.
The problem has been resolved. The default string length is in bytes, use -Dwli.nonxml.strLenInChars=true if the string length is specified in characters
|
|
CR279880
|
Problem: While HTTP control was used, 'java.net.SocketException: Too many open files' exception occurred.
The problem has been resolved. Socket connections are being released in the final block.
|
