I have activated GTF trace to capture RACF trace for RACROUTE calls.
Here is the syslog output:
08.38.28 START GTFRACF.GTF,,,NOPROMPT
08.38.29 STC00107 $HASP373 GTFRACF STARTED
08.38.39 STC00107 AHL121I TRACE OPTION INPUT INDICATED FROM MEMBER
GTFPRM#O OF PDS SYS1.PARMLIB
08.38.39 STC00107 TRACE=USRP
08.38.39 STC00107 USR=(F44),END
08.38.39 STC00107 AHL103I TRACE OPTIONS SELECTED --USR=(F44)
08.38.39 STC00107 AHL080I GTF STORAGE USED FOR GTF DATA:
GTFBLOCK STORAGE 68K BYTES (BLOK= 40K)
PRIVATE STORAGE 1031K BYTES (SIZE= 1024K)
SADMP HISTORY 45K BYTES (SADMP= 40K)
SDUMP HISTORY 45K BYTES (SDUMP= 40K)
ABEND DUMP DATA 0K BYTES (ABDUMP= 0K)
08.38.39 STC00107 AHL031I GTF INITIALIZATION COMPLETE
08.38.59 #SET TRACE(RACROUTE(TYPE(5)) JOBNAME(DANI))
08.38.59 STC00025 IRRH004I (#) RACF SUBSYSTEM SET COMMAND HAS COMPLE
08.39.18 TSU00108 $HASP373 DANI STARTED
08.39.37 stop gtf
08.39.37 STC00107 AHL006I GTF ACKNOWLEDGES STOP COMMAND
08.39.37 STC00107 AHL904I THE FOLLOWING TRACE DATASETS CONTAIN TRACE
08.39.38 STC00107 IEF404I GTFRACF - ENDED - TIME=08.39.38
When I try to format the trace in IPCS the following error is received:
BLS21082I DSNAME('SYS1.TRACE') I/O error
BLS21081I 0AB1,D,SYS00010,READ,OUT OF EXTENT,TTR UNKNOWN,BSAM
AHL10004I Input is not a GTF trace dataset.
AHL10009I No records of the requested type were found.
In TSO the dataset looks like this:
Data Set Name . . . . : SYS1.TRACE
General Data Current Allocation
Management class . . : **None** Allocated tracks . : 100
Storage class . . . : **None** Allocated extents . : 1
Volume serial . . . : USR002
Device type . . . . : 3390
Data class . . . . . : **None** Current Utilization
Organization . . . : PS Used tracks . . . . : 1
Record format . . . : VB Used extents . . . : 1
Record length . . . : 23472
Block size . . . . : 23476
1st extent tracks . : 100
Secondary tracks . : 0
Data set name type : SMS Compressible : NO
Creation date . . . : 2017/03/15 Referenced date . . : 2017/03/15
Expiration date . . : ***None***
Joined: 06 Jun 2008 Posts: 8280 Location: Dubuque, Iowa, USA
In order for IPCS to do anything with the data set, it MUST be defined as RECFM=F,LRECL=4160 or RECFM=FB,LRECL=4160 (although FBS is also allowed, but rarely seen) -- VB,23472,23476 will not allow IPCS to process the GTF trace (or any dump, for that matter).
Joined: 30 Nov 2013 Posts: 618 Location: The Universe
I need to figure out why GTF allocates the file as VB although the DCB in JCL say RECFM=FB.
My guess is the DCB macro within GTF overrode the RECFM=FB in the JCL. This is discussed in JCL reference, and elsewhere.
In terms of personal involvement the last time I used GTF was sometime in the 1980s when I did a CCW trace. I have no memory of what the DCB attributes of the data set were, nor do I recall how I analyzed the data set.
Mr. Sample's statement that RECFM=FBS is rarely seen I do not believe is correct. In my personal involvement it is by far the most common. It appears this is the default when SYSMDUMP is used. I use SYSMDUMP quite a lot in preference to paper SYSUDUMP / SYSABEND dumps as IPCS is far more usable than trying to analyze paper dumps through SDSF.
In this example, the TESTIPCS.DUMP data set used 249 tracks with DCB=(RECFM=FBS,LRECL=4160,BLKSIZE=24960). At first glance this DCB looks pretty sad, but you're not going to do any better. Still, it's better than RECFM=FS, LRECL=4160, BLKSIZE=4160.