PQ59825: THIS APAR ADDRESSES DEFECTS IN WEBSPHERE APPLICATION SERVER V4.0.1 FOR Z/OS AND OS/390. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
![]() |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
![]() APAR status Closed as program error. Error description This APAR addresses defects in WebSphere Application Server V4.0.1 for z/OS and OS/390.Local fix Problem summary **************************************************************** * USERS AFFECTED: All users of WebSphere Application Server * * V4.0.1 for z/OS and OS/390 * **************************************************************** * PROBLEM DESCRIPTION: APAR PQ59825 addresses various problems * * in WebSphere Application Server V4.0.1 * * z/OS and OS/390. * **************************************************************** * RECOMMENDATION: * **************************************************************** APAR PQ59825 addresses the following problems in WebSphere Application Server V4.0.1 for z/OS and OS/390. (MD10950) J2EE Resource Instances are not marked deleted after a refresh. They reappear after a refresh as undeleted. The user chose to delete a J2EE Resource Instance, the Instance was marked deleted as expected. After the user refreshed the the System Management End User Interface (Administration and Operations applications), also known as the SM EUI, the J2EE Resource Instance was no longer marked as deleted, yet it was deleted on the server. (MD12582) Failure to obtain the objectclass attributes during WebSphere LDAP setup results in a generic NULL pointer exception. The generic NULL pointer exception makes it difficult to debug the problem. (MD13249) Running multithreaded applications can lead to intermittent problems getting Runas exceptions. The Container's SecurityCollaborator was not using cookies on a per thread basis. When the collaborator was access by multiple threads, it was possible for an unintialized SecurityCookie to be passed back to the Container. Later, during postInvoke processing the transaction would fail because of this uninitialized cookie. The failure presents itself as: com.ibm.websphere.csi.CSIException: runas problem: no push associated with this pop at com.ibm.ws390.csi.WS390SecurityCollaborator.postInvoke (WS390SecurityCollaborator.java:369) at com.ibm.ejs.container.EJSContainer.postInvoke (EJSContainer.java(Compiled Code)) (MD13281) If an Application is deployed with OS_synchronize specified, or if a connector is deployed with res_auth=container, and the Server EUI setting "Enable Setting OS Thread Identity to RunAs Identity" box isn't checked (i.e., is set to FALSE), the system administrator or user needs to be notified. A user has no way of knowing if syncToThread and server Flag are not set, since no message is currently issued. (MD13431) Bean failures result in generic transaction rolledback exceptions. When an exception occurs in a bean method, the container sometimes masked the root of the problem by just exposing a transaction rolledback exception like: javax.transaction.TransactionRolledbackException: CORBA TRANSACTION_ROLLEDBACK 0 No; nested exception is: org.omg.CORBA.TRANSACTION_ROLLEDBACK: com.ibm.websphere.csi.CSITransactionRolledbackException at java.lang.Throwable.<init>(Throwable.java:92) at java.lang.Exception.<init>(Exception.java:43) at java.io.IOException.<init>(IOException.java:45) at java.rmi.RemoteException.<init>(RemoteException.java:56) at com.ibm.websphere.csi.CSIException.<init> (CSIException.java:31) at com.ibm.websphere.csi.CSITransactionRolledbackException .<init>(CSITransactionRolledbackException.java:26) at com.ibm.ws390.csi.WS390TxStrategy.handleException (WS390TxStrategy.java:161) at com.ibm.ws390.csi.WS390TransactionControl.postInvoke (WS390TransactionControl.java:449) at com.ibm.ejs.container.EJSContainer.postInvoke (EJSContainer.java:2000) The container was only gives detailed information if JRAS tracing was turned on. This results in customers needing to turn tracing on and recreating the problem to determine the real cause of failure. (MD13466) ORB proxy can be leaked in superior nodes of a distributed transaction. When a server enlists in a distributed transaction, a reference to an OTS recovery coordinator is returned. The server that created the recovery coordinator does not release its reference at the end of the transaction. This failure to release the managed object reference causes a storage leak of an ORB proxy per distributed transaction. (MD13560) There is a timeout problem running BBOIRC2 in bootstrap. Customers hit this timeout frequently. Customers need to be provided with instructions on how to solve this timeout issue if it occurs. (MD13569) ABEND0C2/ABENDS0C2 during stop of a Control Region with tracing to buffer active(TRACEBUFFLOC=BUFFER). As sub-threads within the Control Region end, they switch back to problem state. Thread cleanup routines may cut trace records after the switch to problem state. If one of these trace records fills a buffer, then the buffer needs to get written out. The trace writer code tries to switch to key zero from key two. Since we are no longer in supervisor state, the key switch fails with an ABEND0C2. (MD13639) ABEND0C4/ABENDS0C4 in module bbocsess.cpp method Session::post_first_on_pending_queue(). This ABEND results because Communication Session reuse lacks the serialization needed to support freeing a Session while others may still be using it (bbocsess.cpp). (MD13666) Classpath support does not write out Classpath entry into MANIFEST.MF correctly when it exceeds 72 characters in length. Java specifications says that when Classpath entry exceeds 72 characters, it needs to be continued on the next line with initial BLANK. StaticMethodsReflector support does not correctly return EJB methods list when EJB inherits from another EJB. Hence, introducing duplicate EJBObject methods in the returned list and causing EarFileProcessing to run into a NullPointerException during ear file deployment processEARFile (java.lang.NullPointerException from com.ibm.ws390.metadata.archive.EjbJarFile.getAllMethods). PQ51713) If the user installs a J2EE application that contains invalid JNDI names, activates the conversation, and then starts the server, the server will abend with a DC3 abend code (abendDC3), abend reason code 000C0005 and a minor code of C9C21235 during naming registration. The following message will also appear on the console: "BOU0652E NAMING REGISTRATION FAILED FOR SERVER <Server_Name>" This occurs because the DoNamingRegistration method of bbomsbo7.cpp, which is called by the registerHomes method of bboosmoi.cpp, currently throws all naming exceptions at the component level. Any such exception is caught by registerHomes, which then calls a method that brings down the application server. This causes the server to abend, rendering all applications on that server unavailable until the offending application is deleted.Problem conclusion APAR PQ59825 provides the following problem resolution: (MD10950) A new method "displayDatasourceInstanceWithToken" was added to the ISM scripting. This method will be used to display the J2EE Resource Instance with use of a newly added SMDialogToken, which contains the information necessary to display the instances with proper deletion marking. Depending on the ISM scripting version, the SM EUI will dynamically use the old method "displayDatasourceInstance" or if the ISM scripting version is high enough, the new method. (MD12582) A specific message is now produced if the object could not be obtained. This message appears as: Cannot obtain object at specified DN: <customer specific fill-in> (MD13249) Support has been modified to create thread safe access to cookies by moving variable definition into the preinvoke method of SecurityCollaborator. (MD13281) Support has been modified such that additional informational message BBOJ0031I will now be issued to indicate that the SM EUI box "Enable Setting OS Thread Identity RunAs Identity" to enableSyncOSThread is not selected (i.e., set to FALSE). (MD13431) Support has been modified to provide the detailed exception information for bean method failures, even when tracing is turned off. The tracing prints out a more detailed stack trace describing the failure that caused the rollback exception. (MD13446) Support has been modified such that the ORB proxy for the OTS recovery coordinator object is released at end of transaction. (MD13560) The "WebSphere for z/OS Customization" Dialog installation instructions have been updated to address the issue of a timeout problem running BBOIRC2 during bootstrap processing. In the BBOIRC2 step of the Dialog generated installation instruction (BBOINSTR), we tell the customer to look for a timeout. If the time occurs, update the TIMEOUT values temporarily. The added instruction in the BBOIRC2 step looks like this: Check for a return code 0. If the job fails before completing: 1. Check the job log to determine the step at which the failure occurred. 2. Solve the problem that caused the failure. 3. In the job, change the START variable to restart at the failed step. For instance, if the job failed at step 39, change the START variable to read START=39. 4. Resubmit the job. If you get message BBOU0713W and/or abend EC3 rsn-04130001, indicating a time-out, do the following: 1. Update the OTS_DEFAULT_TIMEOUT and OTS_MAXIMUM_TIMEOUT values in the environment file for the &INTINSRV. server instance. The environment file is located in: &CBCONFIG. /controlinfo /envfile /&SYSPLEX. /&INTINSRV./configuration.env Set the values of both environment variables to 3600 or greater. 2. Re-run BBOIRC2, starting at the failed step as explained above. 3. After BBOIRC2 completes successfully, re-set the OTS_DEFAULT_TIMEOUT and OTS_MAXIMUM_TIMEOUT values to 300. (MD13569) Support in module bbortwrt.plx has been modified to not attempt to key change. Key zero is not a requirement when writing out trace records. CTRACEWR documentation states that you must be in Supervisor state or key 0-7. (MD13639) Code has been modified in comm_outbound_locate() (bboocomm.cpp) to not drive freeSession() on a previous SessionHandle if a problem is encountered setting up a subsequent Session. Also, code has been modified to add more serialization in the cleanup of pending ORB_Requests (requests represented in the Pending ORB Request hash table--m_pendingORBRequestHash). (MD13666) Class-Path support has been modified to correctly write out Class-Path entry into MANIFEST.MF correctly when it exceeds 72 characters in length. StaticMethodsReflector support has been modified to correctly return an EJB methods list when EJB inherits from another EJB. ( PQ51713) The DoNamingRegistration method of bbomsbo7.cpp was modified. When it catches a naming registration exception from a component, it will roll back the naming work for that component and then continue on to the next one in the list. The application server will no longer be brought down when the naming regristration for one application fails. The registerComponentNamingRegistrationEntry method of bbomsbo7.cpp was also modified to log a detailed console message when naming registration fails at the component level. This message reads as follows: BBOU0772E NAMING REGISTRATION FAILED FOR COMPONENT <component_name> IN APPLICATION <app_name> ON SERVER <server_name>. THIS APPLICATION WILL NOT BE AVAILABLE UNTIL IT IS REINSTALLED WITH THE NAMING ERROR FIXED FOR THE COMPONENT." The following publication was revised as a result of APAR PQ59825: ________________________________________________________________ WebSphere Application Server V4.0.1 for z/OS and OS/390 Messages and Diagnosis GA22-7837-02 ________________________________________________________________ This APAR requires changes to documentation. NOTE: Periodically, we refresh the documentation on our Web site, so the changes might have been made before you read this text. To access the latest on-line documentation, go to the product library page at: http://www.ibm.com/software/webservers/appserv/ ________________________________________________________________ Chapter 13, pg. 231 (new message) BBOJ0031I WebSphere for z/OS {0} Explanation: An informational message to indicate that system management Server EUI box "Enable Setting OS Thread Identity RunAs Identity" to enableSyncOSThread is not selected (i.e., set to FALSE). User Response: None, if this is the desired mode. If this is NOT the desired setting, check the Server EUI box "Enable Setting OS Thread Identity to RunAs Identity" and rerun the application. ________________________________________________________________ Chapter 13, pg. 360 (new message) BBOU0772E NAMING REGISTRATION FAILED FOR COMPONENT %s IN APPLICATION %s ON SERVER %s. THIS APPLICATION WILL NOT BE AVAILABLE UNTIL IT IS REINSTALLED WITH THE NAMING ERROR FIXED FOR THE COMPONENT. Explanation: Naming registration failed for the specified component. As applications installed on the server do not work properly without being registered in naming, the application to which the component belongs will not function properly. User Response: Delete the application that contains the component with the naming error, then reinstall the application in a new conversation with the corrected naming information for the component. Check accompanying messages in the job log of the System Management server region that processed naming registration for more details. ________________________________________________________________ Chapter 14, pg. 351 (new message) C9C20CEA Explanation: IBM Internal Only User Response: Contact the IBM Support Center ________________________________________________________________ APAR PQ59825 is associated with SERVICE LEVEL W401062 of WebSphere Application Server V4.0.1 for z/OS and OS/390.Temporary fix Comments
APAR is sysrouted FROM one or more of the following: APAR is sysrouted TO one or more of the following: UQ66335 Modules/Macros
|
Document Information |
Product categories: Software > Application Servers >
Distributed Application & Web Servers > WebSphere Application
Server for z/OS
Operating system(s):
Software version: 401
Software edition:
Reference #: PQ59825
IBM Group: Software Group
Modified date: Jun 5, 2002
(C) Copyright IBM Corporation 2000, 2006. All Rights Reserved.