Using the JCL provide by Mr. Engelbrecht, I discovered that the extract worked
as coded. What I also learned was that one of test LPARs was down for an
extended period of time. When the LPAR was IPLd, the latent SMF data was
copied to the input of the extract. Thus, the Type 80 for the first of the
month was encountered in the input phase.
Thank you for this JCL which I will use again and again.
Mr. Mr. Engelbrecht wrote:
>Yes, and since the OP is extracting record type 89, you can do a quick check
>on the output whether the date/time selection works as >advertised.
>Something like this:
//SMF89 EXEC PGM=ICETOOL
//TOOLMSG DD SYSOUT=*
//DFSMSG DD SYSOUT=*
//VLSHCNTL DD *
OPTION COPY,VLSHRT
INCLUDE COND=(6,1,BI,EQ,X'59')
/*
//RAWSMF DD DISP=SHR,DSN=......
//SORTSMF DD DISP=(NEW,DELETE,DELETE),SPACE=(CYL,(50,50,0))
//VREPT DD SYSOUT=*
//TOOLIN DD *
SORT FROM(RAWSMF) TO(SORTSMF) USING(VLSH)
DISPLAY FROM(SORTSMF) LIST(VREPT) -
HEADER('SMFTYP') ON(6,1,BI) -
HEADER('DATE') ON(11,4,DT1,E'9999/99/99') -
HEADER('TIME') ON(7,4,TM1,E'99:99:99') -
HEADER('SYSTEM') ON(15,4,CH)
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
