Page 1 of 2

SMF record issue

PostPosted: Tue Oct 26, 2010 8:46 pm
by jaggz
Hi,

We were planning for SMF plain performance report. While running the JCL i got a Severity Code of 016 in the statement. Spool says that SMFIN offloaded records were not found. Though we have stored the SMFREC dataset in LPAR.

SYS1.ZOSB.SMREC - >>>>>>>>>>>>>> this record was not executing after the EXTR exec . The condition code was = 016. Usually the SMFREC(offloaded) Would be a generic dataset. To check in 3.4 option, i was not able to open this dataset = SYS1.ZOSB.SMFREC .

Below is JCL which we ran and got the Severity Code of 016.


//#xxxxx JOB (????,????),MSGLEVEL=1,MSGCLASS=O,NOTIFY=????????
//*
//* THIS IS JCL TO RUN THE CP2000 EXTRACT PROGRAM
//* SEE THE USER'S GUIDE FOR COMPLETE DESCRIPTION OF
//* PARAMETERS AND DD STATEMENTS
//*
//* FIND THE USER'S GUIDE AT
//* W3.IBM.COM/SUPPORT/AMERICAS/WSC/CPSPRODUCTS.HTML
//* (NOTE THE URL ABOVE MUST BE IN LOWER CASE.)
//*
//EXTR EXEC PGM=LOADER
//* PRINT DATA SET FOR MESSAGES FROM YOUR RUN
//PRINT001 DD SYSOUT=*
//* INPUT SMF DATASET
//SMFIN DD DISP=SHR,DSN=SYS1.ZOSB.SMREC - >>>>>>>>>>>>>> this record was not executing after the EXTR exec . The condition code was = 016
//* OUTPUT DATASET FOR EDFI FILE
//EDF001 DD DISP=SHR,DSN=PATTRSN.CPSTOOLS.JCL(EDFI)
//* OUTPUT DATASET FOR WORKLOAD GROUPING DATASET
//PGN001 DD DISP=SHR,DSN=PATTRSN.CPSTOOLS.JCL(PGNMAP)
//* OUTPUT DATASET FOR AUTO BCU MAP
//BCU001 DD DISP=SHR,DSN=PATTRSN.CPSTOOLS.JCL(BCUMAP)
//* COMPAT MODE PROCESSING OF IEAIPS TO GENERATE PGNMAP
//*IPS001 DD DSN=SYS1.PARMLIB(IEAIPS00),DISP=SHR
//* INPUT DATASET FOR IOCP PROCESSING (FICON ANALYSIS)
//*IOCP001 DD DISP=SHR,DSN=PATTRSN.CPSTOOLS.JCL(IOCP)
//* OUTPUT DATASET FOR DISK MAGIC OUTPUT
//*MAGIC001 DD DISP=SHR,DSN=PATTRSN.CPSTOOLS.JCL(MAGIC)
//* DDS FOR OPTIONAL INTERNAL SORT OF SMF RECORDS
//*SORTMSGS DD SYSOUT=*
//*SYSOUT DD SYSOUT=*
//*SORTWK01 DD UNIT=SYSDA,SPACE=(CYL,(200))
//*SORTWK02 DD UNIT=SYSDA,SPACE=(CYL,(200))
//*SORTWK03 DD UNIT=SYSDA,SPACE=(CYL,(200))
//* # OUTPUT DATASET FOR TYPE30 RECORDS
//* NOTE: DCB INFORMATION FOR THESE DATASETS IS SUBJECT TO CHANGE.
//* SO IT IS BEST TO LET THE PROGRAM ALLOCATE THEM.
//*TYP30001 DD SPACE=(CYL,(20,10),RLSE),DISP=(,CATLG),UNIT=SYSDA,
//* DSN=PATTRSN.TYPE30.DATA
//* # OUTPUT DATASET FOR SAVE= PARAMETER (SAVE CP2K'S SMF RECORDS)
//*SMFSAVE DD SPACE=(CYL,(20,10),RLSE),DISP=(,CATLG),UNIT=SYSDA,
//* BLKSIZE=8192,DSN=PATTRSN.SMF.SAVE
//* # OPTIONAL USER DUMP DATASET
//*SYSUDUMP DD SYSOUT=*
//SYSIN001 DD *
ENT='zCP3000 INC.'
SYSID=P84M
DURATION=1
DATE=(09/13/94)
TIME=(08-12)
PGN=GOAL
BCU=AUTO READ SMF AND GENERATE AND PROCESS BCUMAP
/*
//* MESSAGES FROM THE MVS LOADER
//SYSLOUT DD SYSOUT=*
//* THE OBJECT MODULE OF THE PROGRAM
//SYSLIN DD DISP=SHR,DSN=PATTRSN.CPSTOOLS.JCL(ZOBJEXTR)

Re: SMF record issue

PostPosted: Tue Oct 26, 2010 11:15 pm
by dick scherrer
Hello,

You need to post all of the diagnostic information generated by the run. . .

Somewhere, there is an explanation of the condition code 0016.

DD statements do not cause condition codes - EXEC statements cause condition codes.

Re: SMF record issue

PostPosted: Tue Oct 26, 2010 11:36 pm
by enrico-sorichetti
since You are using CP2000 which is an IBM internal only tool
there are two alternatives ...

1) You are an IBM employee
2) You are using CP2000 after an agreement with IBM

in both cases issues on the CP2000 usage should be addressed to IBM!
in case (1) Your tutor
in case (2) the SE responsible for technical support of Your site or as an alternative to your salesdroid

Re: SMF record issue

PostPosted: Wed Oct 27, 2010 6:29 am
by steve-myers
Why couldn't you open SYS1.ZOSB.SMFREC? Most SMF data sets are RECFM=VBS, which ISPF can't handle. If that's the reason, well ... Even if ISPF could open the data set, most likely you couldn't do very much with the data; it's mostly binary data.

The comment you added to the SMFIN DD statement really doesn't make any sense.

Specifying PDS data sets for output as you have done is extremely risky. If the program opens and closes them one at a time it's safe, othwerwise data will likely be corrupted.

DATE=(09/13/94) doesn't make much sense. I know nothing about CP2000 beyond what I have read here, but if this parameter is specifying a date as selection criteria for SMF records to process, I suspect there are no records with that date in your input.

Re: SMF record issue

PostPosted: Wed Oct 27, 2010 7:51 am
by jaggz
I very well know that the SMF record formats are VB . Problem is here that the SMFIN Input dataset was not readable by us(Our own racfid) in the program not in real sense of opening(Got to know via the Spool). Since we executed this JCL from our own ID.Is it a problem with our ID that it does'nt have the previlege to Read during execution or else can make the Job card to be specific to $IBMUSER..

Re: SMF record issue

PostPosted: Wed Oct 27, 2010 7:59 am
by dick scherrer
Hello,

I very well know that the SMF record formats are VB .
Either you don't know very well or you made a typo. . . :)

These are not VB - they are VBS.

Do you now have the output you wanted?

Re: SMF record issue

PostPosted: Wed Oct 27, 2010 11:13 am
by jaggz
apology for the wrong representation.


The only change i made here was SORT =NO.... the condition code was 012. A small thought should i be changing the record format from VB to FB ?????

The output is

CP2KT30M: NO T30M001 PROCESSING REQUESTED
CP2KES74: WARNING- BCU MAP FOR VZOTLGR SPECIFIED 2105-800
WHEN LCU DATA HAD 2105
4 ,EX72: R723CICT INVALID (WIORESPV) SEE APAR OA04856 INT 60. 09/08/10 11:00
CP2KES72: PGNDESC=STCTASKS.OMVS.1
4 ,EX72: R723CICT INVALID (WIORESPV) SEE APAR OA04856 INT 184. 09/13/10 15:00
CP2KES72: PGNDESC=STCTASKS.OMVS.1
4 ,EX72: R723CICT INVALID (WIORESPV) SEE APAR OA04856 INT 232. 09/15/10 15:00
CP2KES72: PGNDESC=STCTASKS.OMVS.1
4 ,EX72: R723CICT INVALID (WIORESPV) SEE APAR OA04856 INT 255. 09/16/10 14:00
CP2KES72: PGNDESC=STCTASKS.OMVS.1
4 ,EX72: R723CICT INVALID (WIORESPV) SEE APAR OA04856 INT 272. 09/17/10 07:00

CP2KES72: PGNDESC=STCTASKS.OMVS.1
8 ,ES70: DURATION: DATA NOT EQUAL TO PARAMETER IN INT 281. 09/20/10 14:00
12 ,EXTR: MISSING TYPE 70 RECORD AT 09/18/10 AT 10:30. SORT OR EXCLUDE INTERVAL
8 ,ES70: DURATION: DATA NOT EQUAL TO PARAMETER IN INT 352. 09/20/10 14:30
4 ,EX72: R723CICT INVALID (WIORESPV) SEE APAR OA04856 INT 455. 09/24/10 21:00
CP2KES72: PGNDESC=STCTASKS.OMVS.1
4 ,EX72: R723CICT INVALID (WIORESPV) SEE APAR OA04856 INT 546. 09/28/10 16:00
CP2KES72: PGNDESC=STCTASKS.OMVS.1
8 ,ES70: DURATION: DATA NOT EQUAL TO PARAMETER IN INT 592. 09/30/10 14:00
CP2KEXTR: 21572 DUPLICATE SMF RECORDS ELIMINATED
CP2KEXTR: SMF RECORDS NOT IN TIME SEQUENCE.
4 ,EXTR: MORE SATISFACTORY RESULTS MAY BE OBTAINED BY SORTING: SORT=YES.
CP2KES42: TOTAL I/O RATE PROCESSED= 10.5
CP2KES42: INFO MSG: MAX TYPE 42 INTERVAL: 85354 MINUTES JOB IXGLOGR
CP2KES72: REPORT PERFORMANCE GROUPS FOUND:
RCB
RIWEB
RMQ
CP2KEXTR: PROCESSING COMPLETE

Re: SMF record issue

PostPosted: Wed Oct 27, 2010 11:22 am
by enrico-sorichetti
again...
why do You refuse to ask IBM?
it would be certainly the fastest way to get out of Your issue

Re: SMF record issue

PostPosted: Wed Oct 27, 2010 11:45 am
by jaggz
Sure Enrico I will ask the IBMERS and Let you know once we are onto track. :)

Re: SMF record issue

PostPosted: Wed Oct 27, 2010 12:48 pm
by jaggz
Its resolved now. It was Just the Some SMF record intervals were missing, which were not important for our analysis.