Preventive Service Planning
Experiencing PDS issues in my agent and / or Tivoli Management Server (TMS), how do I debug these issues.
This document will assist you in troubleshooting a PDS issue in z/OS. This addresses V622, V623 and V630 of IBM Tivoli Monitoring (ITM).
Updated maintenance for PDS is in PTF UA62291 (ITM V622 FP6) and in ITM V623 base.
There were ESTAE changes to PDS to perform a graceful shutdown should a problem with PDS arise. All efforts will be made to write the data from the PDS buffers to the PDS files. Also, PDS dumps will contain information that will help support isolate the problem, if possible.
APAR OA41236 has been opened to produce better error messages prior to abending.
Some of the error messages you will see in either the RKPDOUT or the RKPDLOG output files:
- KPDXIO: File not eligible for writing
- LSCX041 **** ERROR ****
ABEND occurred in line 187 of @@nnnnn(ITBL) ,offset 000266
Program terminated by operating system. ABEND code = S0C4
- LSCX010 **** ERROR ****
ABEND occurred in ASRTFL called from line 518 of _DYNAMN(MANE) ,
Program terminated by call to abort. ABEND code = U1210
- KPDIFIL: Invalid status information detected in file hiLevel.Qualifier.table name
(group group-name).........The file had not been closed properly.
- ABEND U1213 in KPDMANE
Note: Do not delete and reallocate the PDS files unless it is the last resort.
- Verify ITM V622 is at FP6 or higher.
- Send in the SVCDUMP title:
KPDMANE: Persistent DataStore ESTAE Dump followed by a message in the RKLVLOG:
- Send the SVCDUMP to IBM support along with the complete started task log. This is either the agent or TMS log depending on where PDS abended.
- If you are receiving message "KPDMN004 PDS IS INACTIVE", history collection is stopped. Currently there is no command to activate PDS once it has been set to an inactive state. A recycle of the address space is required to restart history collection.
- PDS should have automatically recovered the PDS files. If not, issue the following command:
- F address-space,KPDCMD RECOVER FILE=DSN:datasetname (where datasetname is the full-qualified name of the file, without quotes)
- If PDS continues to abend, you can execute a batch job to try to recover the data. The batch jobs can be found in &rhilev.&rte.RKANSAMU:
- KPDEXPTJ - Batch job to export a persistent datastore
- KPDRESTJ - Batch job to restore a persistent datastore from a file that was created by KPDEXPTJ.
- KPDXTRAJ - Batch job to extract table data from a persistent datastore dataset.
- If all else fails, you can scratch and reallocate the PDS files. In PARMGEN the WKANSAMU dataset contains standalone jobs to reallocate the PDS files if you do not want to run the composite KCIJPALO RTE build job. In ICAT, you can execute the #P and #Q jobs after deleting the PDS files.
Enhancements are currently being made to the Persistent Datastore error processing for the following issues:
- To the PDS ESTAE routines
- To terminate gracefully should an abnormal termination occur
|Systems and Asset Management||Tivoli OMEGAMON XE for CICS on z/OS||4.2.0, 4.1.0, 5.1.0|
|Systems and Asset Management||Tivoli OMEGAMON XE on z/OS||4.1.0, 4.2, 5.1.0|
|Systems and Asset Management||Tivoli OMEGAMON XE for CICS TG on z/OS||4.2.0, 5.1.0|
|Systems and Asset Management||Tivoli OMEGAMON XE for DB2 Performance Expert for z/OS||4.2.0, 4.1.0, 5.1.0, 5.1.1|
|Systems and Asset Management||Tivoli OMEGAMON XE for DB2 Performance Monitor for z/OS||5.1.0, 5.1.1|
|Systems and Asset Management||Tivoli OMEGAMON XE for IMS on z/OS||4.2.0, 4.1.0, 5.1.0|
|Systems and Asset Management||Tivoli OMEGAMON XE for Mainframe Networks||4.2.0, 4.1.0, 5.1.0|
|Storage Management||Tivoli OMEGAMON XE for Storage on z/OS||4.2.0, 4.1.0, 5.1.0|
|Systems and Asset Management||Tivoli OMEGAMON XE for Messaging for z/OS||6.0.1, 6.0.0, 7.0, 7.0.1, 7.1.0|