PQ88979: Control Region abends with 0c4 when restarting a servant region to replace a previously abended one (ie. due to timeout) | |||||||||||||||||||||||||||||||||||||||||||
![]() |
|||||||||||||||||||||||||||||||||||||||||||
![]() APAR status Closed as program error. Error description When a servant region for some reason abends and is terminated (for example due to a timeout), the recovery code should clean up any requests in process or if early enough, transfer them to a newly started servant region to continue request execution. There are 2 recovery processes that run asynchronously. There is a small window where one recovery procedure can transfer a request to a newly started servant region. The new servant region will start to generate a response to the request. In the meantime, the 2nd recovery process will try to clean up the request thinking it is in the 1st servant region. At that time the response to to the client fails as ORBR was cleaned up in error causing control region to abend with 0C4 abend code and RC 11. We were in BBOO_ORBR_Protected_Info::buildORBR_Protected Info(void*,SEQUENCE_octet*) when trying to do a load from address at register 2, but register to contained 000A0000. _ The treceback of the failing TCB: CEEO1PTQ CEEKDUMP csnap RasSignalHandler2 CEEPGTFN __zerros CEEHDSP BBOO_ORBR_Protected_Info::buildORBR_Protected_Info(void*,SEQ Exception BBOO_ORBR_Protected_Info::setSecurityExecutionEnv(ORBR_Prote SecurityManager::setSecurityExecutionEnv(SEQUENCE_octet*) ORB_Request::reconnectToCurrentThread() ORB_Request::startOfRequestOnTask() ACR_ExecutionThread::RemoveAndProcessWork(ThreadCleanUp*)Local fix Problem summary **************************************************************** * USERS AFFECTED: All users of WebSphere Application Server * * V5.0 for z/OS * **************************************************************** * PROBLEM DESCRIPTION: ABENDS0C4/ABEND0C4 in * * BBOO_ORBR_Protected_Info:: * * buildORBR_Protected_Info in a * * Websphere Application Controller * **************************************************************** * RECOMMENDATION: * **************************************************************** An ORB_Request was cleaned up while still in use resulting in an ABEND 0C4 during outbound response processing within a controller. This may occur if a servant region is abnormally terminated (ex., due to a timeout): the recovery code should cleanup any requests in-flight or if possible, transfer them to another servant region for execution. There are 2 recovery processes that run asynchronously: the "SR Terminated" recovery routine and the "SR Task" recovery routine. There is a small window where the "SR Terminated" recovery routine can transfer a request to another servant region while the "SR Task" recovery routine will start to generate a response to the client and cleanup the request. After the requeued request is processed in another servant region the response will be driven back through the Controller region. At this time the second response to the client fails as the request (ORB_Request) was already cleaned up. In this case the error causing the controller region to ABEND was an ABENDS0C4.Problem conclusion Code has been modified to serialize the "SR Terminated" and "SR Task" recovery routines. The following publication was revised as a result of APAR PQ88979: ________________________________________________________________ WebSphere Application Server V5 for z/OS Messages and Codes GA22-7915-01 _______________________________________________________________ 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: www.ibm.com/software/webservers/appserv/zos_os390/library.html ________________________________________________________________ Chapter 3, pg. 125 (new message) Message identifier - C9C2130A Explanation: IBM Internal Only User Response: Contact the IBM Support Center. ________________________________________________________________ Chapter 3, pg. 125 (new message) Message identifier - C9C21315 Explanation: IBM Internal Only User Response: Contact the IBM Support Center. ________________________________________________________________ Chapter 3, pg. 125 (new message) Message identifier - C9C2132C Explanation: IBM Internal Only User Response: Contact the IBM Support Center. ________________________________________________________________ Chapter 3, pg. 125 (new message) Message identifier - C9C2132D Explanation: IBM Internal Only User Response: Contact the IBM Support Center. ________________________________________________________________ APAR PQ88979 is associated with SERVICE LEVEL W502013 of WebSphere Application Server V5.0 for z/OS.Temporary fix Comments
APAR is sysrouted FROM one or more of the following: PQ88977 APAR is sysrouted TO one or more of the following: PQ89478 Modules/Macros
Publications Referenced
|
Document Information |
Current web document: swg1PQ88979.html
Product categories: Software > Application Servers >
Distributed Application & Web Servers > WebSphere Application
Server for z/OS
Operating system(s):
Software version: 500
Software edition:
Reference #: PQ88979
IBM Group: Software Group
Modified date: Aug 4, 2004
(C) Copyright IBM Corporation 2000, 2009. All Rights Reserved.