PQ75519: WebSphere for z/OS sometimes can't start new server regions with BBOO_WORKLOAD_PROFILE=IOBOUND or BBOO_WORKLOAD_PROFILE=CPUBOUND | |||||||||||||||||||||||||||||||||||||||||||
![]() |
|||||||||||||||||||||||||||||||||||||||||||
![]() APAR status Closed as program error. Error description WebSphere for z/OS is unable to start additional server regions when the server's environment variable BBOO_WORKLOAD_PROFILE is set to IOBOUND or CPUBOUND when the number of CPUs active on the processor has been changed since the server itself was started. The following messages are received: BBOU0041E Function IWMCONN failed with RC=8, REASON=1104085C BBOU0006E CB SERIES SERVER PW2FONEA ENDED ABNORMALLY,REASON=C9C21031 The syslog says *IWM032I INTERNAL STOP FOR PW2FONE COMPLETED The APPLENV for the server is stopped. Resuming the APPLENVLocal fix does not c ear the problem. The only way to clear the problem is to stop the control region and restart it. Each server region that is started must have the same number of WLM worker threads. The number of worker threads is computed based on the number of currently active CPUs if BBOO_WORKLOAD_PROFILE is set to IOBOUND or CPUBOUND. If the number of active CPUs changes over time, it will be impossible to create new server instances because the requested number of WLM worker threads will be different than the number established for the server on that system. LOCAL FIX:ン Use BBOO_WORKLOAD_PROFILE values of NORMAL or LONGWAIT, or remove BBOO_WORKLOAD_PROFILE completely from current.env.Problem summary **************************************************************** * USERS AFFECTED: All users of WebSphere Application Server * * V5.0 for z/OS * **************************************************************** * PROBLEM DESCRIPTION: Additional server regions fail to start * * when the server's environment variable * * BBOO_WORKLOAD_PROFILE is set to * * IOBOUND or CPUBOUND when the number of * * CPUs active on the processor has been * * changed since the server itself was * * started. The following messages are * * received: * * BBOU0041E Function IWMCONN failed with * * RC=8, REASON=1104085C * * BBOU0006E CB SERIES SERVER * * xxxxxxxx ENDED ABNORMALLY, * * REASON=C9C21031 * **************************************************************** * RECOMMENDATION: * **************************************************************** When a server region comes up it determines the number of WLM worker threads based on the number of active CPUs if the server's environment variable BBOO_WORKLOAD_PROFILE is set to IOBOUND or CPUBOUND. If number of active CPUs is changed since the server itself was started, WebSphere for z/OS is unable to start additional server regions because the number of WLM worker threads will be different than the number established for the server on that system. The number of worker threads must be consistent.Problem conclusion Code was modified such that the initial number of threads is calculated once based on the number of CPUs when the control region initializes. Even though number of CPUs might change in between server regions startups, they all get the same value from the control region. The number of threads remains same for all server regions. APAR PQ75519 is associated with SERVICE LEVEL W500103 of WebSphere Application Server V5.0 for z/OS.Temporary fix Comments
APAR is sysrouted FROM one or more of the following: PQ74824 APAR is sysrouted TO one or more of the following: Modules/Macros
Publications Referenced
|
Document Information |
Current web document: swg1PQ75519.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 #: PQ75519
IBM Group: Software Group
Modified date: Sep 5, 2003
(C) Copyright IBM Corporation 2000, 2009. All Rights Reserved.