APAR status
Closed as fixed if next.
Error description
Problem 1: SYSTEM COMPLETION CODE=A03 TIME=23.29.27 SEQ=32713 CPU=0000 ASID=0014 PSW AT TIME OF ERROR 070C2000 84C68F16 ILC 2 INTC 0D NO ACTIVE MODULE FOUND NAME=UNKNOWN DATA AT PSW 04C68F10 - 5810800C 0A0D58E0 02FC58D0 GR 0: 7F53F000 1: 80A03000 2: 008EADF8 3: 04C5F01F 4: 008FF530 5: 7F53E600 6: 008FF530 7: 008FF310 8: 04C6A118 9: 008EAE04 A: 00000000 B: 00F90600 C: 008EADF8 D: 008EAE64 E: 84C61B18 F: 84C68E98 END OF SYMPTOM DUMP 08246 23:29:27.68 JOB07720 00000094 IEF450I T70X3834 S010 PA8044 - ABEND=S222 U0000 REASON=00000000 084 084 00000094 TIME=23.29.27 08246 23:29:28.58 00000090 IGD022I THE SMS ADDRESS SPACE HAS FAILED AND IS RESTARTING RIDS/IKJEFT09#L RIDS/#UNKNOWN AB/S013E REGS/0E086 REGS/0A156 SYMPTOM DESCRIPTION ------- ----------- RIDS/IKJEFT09#L LOAD MODULE NAME: IKJEFT09 RIDS/#UNKNOWN CSECT NAME: #UNKNOWN AB/S013E SYSTEM ABEND CODE: 013E REGS/0E086 REGISTER/PSW DIFFERENCE FOR R0E: 086 REGS/0A156 REGISTER/PSW DIFFERENCE FOR R0A: 156 JOBNAME: IEESYSAS SYSTEM NAME: PROD ERRORID: SEQ=32712 CPU=0000 ASID=0014 TIME=23:29:27.7 When a job either creates, deletes, or alters a dataset it will use SMS to do that. So when a job is canceled through either a program or TSO user it results in a ABEND222 (The operator, or an authorized time sharing option extensions (TSO/E) user, canceled the job without requesting a dump.). So for each 222 abend there will be a resulting 13E (The task that created a subtask issued a DETACH macro for that subtask, specifying STAE=NO, before the subtask ended.) As 13E abend represents a FORCE and a DETACH of a SUBTASK. The FORCE resulted an SMS FRR (Functional Recovery Routine) getting control which like many FRRs calls IGDERDMP to take a DUMP. Because of the cancels within a SMS task, i.e. allocation, extending, or other dataset processing this consequently brings down the SMS address space. Problem 2: After issuing message IGD0221I, SMS subsystem initialization routine IGDSSI00 gets control. IGDSSI00 links to IEEMB881 to start the SMS address space and there are no specific attributes that are requested (in particular, the PRIV attribute is not set). IEEMB881 issues a SYSEVENT EASINIT and if the PRIV attribute is not set and if there is no classification rule for SMS, then the newly created SMS address space will be assigned the default service class for subsystem STC. Further on in initialization processing, IEFSD263 will issue a SYSEVENT INITATT and the parameter list contains the System Task and Privileged attributes from the PPT. In the IBM supplied PPT, SMS is marked as a System Task. During SYSEVENT INTATT processing, the address space is classified again and since the System Task attribute was specified, it will default into service class SYSSTC.
Local fix
Problem 1: The only way to avoid this problem is to prevent the multiple job cancels (S222, user abends). If this is not possible then ignore the symptoms and use DAE dump suppression to suppress the dumps. NOTE: DAE dump suppression only suppresses the dumps and will not prevent the symptoms of the problems. Problem 2: Ensure SMS has a high priority classification rule in the WLM. Add a WLM classification rule for SMS and assign it service class SYSSTC. This will prevent a long delay on the restart of SMS.
Problem summary
**************************************************************** * USERS AFFECTED: All users of DFSMS 1K0 or higher. * **************************************************************** * PROBLEM DESCRIPTION: SMS issues the IGD023I message that * * indicates that the SMS address space * * has restarted. This is required because * * owing to asynchronous abends, a * * critical error has been detected in the * * SMS address space. * **************************************************************** * RECOMMENDATION: * **************************************************************** Currently, when this happens, a DUMP is not taken.
Problem conclusion
Temporary fix
Comments
CONTROL NUMBER: KFI0537 This APAR is being closed as FIN with the concurrence of the submitting customer. A solution to this problem will be delivered in a future DFSMS release, if there is one, within the next 18 to 24 months. This problem is fixed in z/OS DFSMS V1R11
APAR Information
APAR number
OA26435
Reported component name
STORAGE MGMT SU
Reported component ID
5695DF101
Reported release
1K0
Status
CLOSED FIN
PE
NoPE
HIPER
NoHIPER
Special Attention
NoSpecatt / Xsystem
Submitted date
2008-09-08
Closed date
2008-10-16
Last modified date
2013-04-10
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
Fix information
Applicable component levels
R1K0 PSN
UP
[{"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Product":{"code":"SG19M","label":"APARs - z\/OS environment"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"1K0","Edition":"","Line of Business":{"code":"","label":""}}]
Document Information
Modified date:
10 April 2013