How to: Collect diagnostic information using PDCOLLECT

Edit online

Use the Problem Determination Data Collection (PDCOLLECT) tool to collect diagnostic information. This topic explains how to perform this task using PARMGEN.

Before you begin

Note: If you use Configuration Manager as your configuration tool, see Collecting diagnostic data using PDCOLLECT.

Use the WCONFIG(KCI$SPDC) KCIPDCOL imbed of the KCIJPCOL PDCOLLECT job to collect diagnostic information from the machine on which the product address spaces are executed.

To use this tool, you must install SDSF and are licensed to use it. If you do not have SDSF, you can copy the complete address space logs, including the JES output, RKLVLOG, RKPDOUT, RKPDLOG, SYSPRINT, and so on, into a data set. The DCB information for the output data set can be as follows:

Organization: PS  
Record format: VB 
Record length: 240
Block size: 27998

Use the following format:

//* ---------------- BEGIN - USER SECTION: OVERRIDE -------------- *
//KCIPDCOL EXEC PGM=IKJEFT1A,REGION=0M,
//         DYNAMNBR=99,COND=EVEN  
//SYSTSPRT DD SYSOUT=*  
//SYSPRINT DD SYSOUT=*  
//SYSUDUMP DD SYSOUT=*  
//SYSTSIN DD *
  EX '&gbl_target_hilev.TKANCMD(KCIPDCOL)' +  
     'JOBOUT(data_set_name) +
      SYS(&rte_name) +
      RHILEV(&rte_hilev) + 
      RVHILEV(&rte_vsam_hilev) +  
      BASEHLEV(&rte_hilev.&rte_name.R) + 
      PGNJCL(&gbl_user_jcl) +  
      PGNRTE(&rte_plib_hilev.&rte_name.WCONFIG) +
      INCLQA1(Y)'  
/*  
//* ---------------- END -USER SECTION: OVERRIDE -------------- *

About this task

The PARMGEN KCIJPCOL job uses the problem determination data collection tool KCIPDCOL to collect data that includes the following aspects:

System configuration
Runtime environment (RTE) libraries, including configuration files
Network information, such as output from the netstat utility
Self-describing agent (SDA) information:
- Results of ls -al for your rte/kds/support and rte/kds/support/TEMS
- Contents for spawn_cmd.sh and spawn_cmd.trs
- Results of java – version
Output from the specified job, including logs for the requested started task

Before you submit the $PARSE* "Create runtime members and jobs", review the RTE's WCONFIG(KCI$SPDC) override imbed to KCIJPCOL to decide whether you have additional collection options you want to enable.

Note: The WCONFIG(KCI$SPDC) imbed might not be the latest version. Check the RTE's IKANSAMU(KCI$SPDC) template for changes to the imbed if you want to implement it in your user copy in WCONFIG(KCI$SPDC), and then rerun the $PARSESM job (which recreates the runtime members from IKANSAMU to WKANSAMU) and KCIJPW2R job to refresh the KCIJPCOL job.

Procedure

Define the RTE that you want to configure.
1. Start the installation and configuration tool by running the command EX 'gbl_target_hilev.TKANCUS' where gbl_target_hilev is the high-level qualifier for the SMP/E target libraries where the products are installed.
  The Welcome screen for the z/OS Installation and Configuration Tools is displayed.
2. Select 3 (Configuration Workflow), and press Enter.
3. Specify the information of the RTE that you want to configure including GBL_USER_JCL, RTE_PLIB_HILEV, and RTE_NAME, and press Enter.
  PARAMETER GENERATOR (PARMGEN) Workflow - Primary Option Menu is displayed.
Type U or UTIL in the Option line and press Enter to go to Utilities panel.
Select option 32 to configure the KCIJPCOL job.
- Default parameters are provided in the USER SECTION. Add your advanced PDCOLLECT parameters, if any.
- The JOB(NAME(IBM*) parameter of KCIJPCOL is set to collect data based on the global RTE started task prefix value of RTE_STC_PREFIX=IBM. If you use started task names different from NAME(IBM*), modify the NAME() parameter accordingly. Each KCIJPCOL submission collects data for STCs matching NAME(IBM*) convention.
- You can choose whether to include the HUB TEMS &rte_plib_hilev.&rte_name.RKDS* VSAM datasets that are referenced in the QA1* DDNAMEs in the IBMDS TEMS started task. The default value is Y.
- For %&var% parameter variables, do not modify actual PARMGEN profile parameters that are referenced as RTE_HILEV_BASEHLEV, RTE_STC_PREFIX, GBL_USER_JCL, or RTE_PLIB_HILEV.RTE_NAME. PARMGEN's $PARSE* job automatically substitutes these variables with actual customization values that are supplied in the LPAR RTE profile (WCONFIG(RTE_NAME)). WCONFIG(KCI$SPDC) override imbed is embedded into xKANSAMU(KCIJPCOL) job by the $PARSE* "Create runtime and members" job.
Assume that an environment has the following settings:
- RTE_NAME=$DEFAULT
- RTE_TYPE=SHARING RTE (allocates its data sets as TDITN.FTU.RTENVSM.$DEFAULT.RKANPARU, TDITN.FTU.RTENVSM.$DEFAULT.RKANDATV, TDITN.FTU.RTEVSAM.$DEFAULT.RKDSSITF, and so on)
- This LPAR sharing RTE shares a read-only set of base libraries that are called 'TDITN.FTU.RTEBASEA.RK*' (that is, TDITN.FTU.RTEBASEA.RKANCMD, TDITN.FTU.RTEBASEA.RKANMOD, TDITN.FTU.RTEBASEA.RKANMODL, and so on)
- This $DEFAULT sharing RTE uses a different HLQ for its PARMGEN work data sets (TDITN.FTU.PGNWORK.$DEFAULT.W*, that is, TDITN.FTU.PGNWORK.$DEFAULT.WCONFIG, TDITN.FTU.PGNWORK.$DEFAULT.WKANPARU, and so on)
The settings of the KCIJPCOL job are as follows:
```
//* ---------------- BEGIN - USER SECTION: OVERRIDE -------------- *
//KCIPDCOL EXEC PGM=IKJEFT1A,REGION=0M,
           DYNAMNBR=99,COND=EVEN  
//SYSTSPRT DD SYSOUT=*
//SYSPRINT DD SYSOUT=*  
//SYSUDUMP DD SYSOUT=*  
//SYSTSIN DD *  
  EX 'TDITN.FTU.RTEBASEA.RKANCMD(KCIPDCOL)' +  
     'JOB(NAME(IBM*) NO(STC*)) +  
      SYS($DEFAULT) +  
      RHILEV(TDITN.FTU.RTENVSM) +   
      RVHILEV(TDITN.FTU.RTEVSAM) + 
      BASEHLEV(TDITN.FTU.RTEBASEA.R) +  
      PGNJCL(TDITNT.COMMON.PARMGEN.GLOBAL.JCL) + 
      PGNRTE(TDITN.FTU.PGNWORK.$DEFAULT.WCONFIG) +  
      INCLQA1(Y)'
/* 
//* ---------------- END - USER SECTION: OVERRIDE -------------- *
//****************************************************************** 
```
Submit the job.

Results

The file &SYSUID.KCIPDCOL.PDCOLPDS.TRS is generated. You can FTP the output data set to IBM Support Center.