Planet MVS Flower Box

• Frequently Asked Questions
• My current Flower Box
• IBM's sample IEFACTRT
• Contributed "Flower Box" output
• Disclaimers

What is a "Flower Box"?

A Flower Box, as it relates to the OS/390 (MVS) operating system, is the combined job and step end informational records which are written to the SYSMSG JES dataset by the SMF exit IEFACTRT. They are sometimes surrounded by a box of astericks or dashes. Somebody along the line gave them the cute nick name Flower Box. Here's an example with the "flower box" in red:

******************************************************* * STEP END STATISTICS: STEP JS01 RC=0015 * * I/O: SYSUT1 1283 SYSUT2 1283 ******************************************************* IEF375I JOB/IBMUSERT/START 1999182.1022 IEF376I JOB/IBMUSERT/STOP 1999182.1027 CPU 0MIN 48.

Where do I find them?"

Here's a JES2/SDSF example:

• Select a job's output via the "?" line command in the hold or output queue.
• You will see a list of JES2 sysout groups like:
  
  

  SDSF JOB DATA SET DISPLAY - JOB IBMUSERG (JOB00050) COMMAND INPUT ===> PREFIX=* DEST=(ALL) OWNER=* NP DDNAME STEPNAME PROCSTEP DSID OWNER C DEST JESMSGLG JES2 2 IBMUSER X LOCAL JESJCL JES2 3 IBMUSER X LOCAL JESYSMSG JES2 4 IBMUSER X LOCAL HCDMLOG IOCP0 GO 102 IBMUSER X LOCAL

  
  
• Browse the one called "JESYSMSG"
• Look for output giving statistics for the step and job end. The will be by these messages:

  IEF373I STEP/stepname/START ...
  IEF374I STEP/stepname/STOP ...
  ...
  IEF375I  JOB/jobname /START ...
  IEF376I  JOB/jobname /STOP ...
  

It's possible that your shop doesn't create a "Flower Box" from the IEFACTRT SMF exit. If this is so, you may only see only IEF and IGD prefixed messages in your SYSMSG.

Shouldn't the IBMUSER userid be disabled?

I don't really use the userid IBMUSER. I only use it in examples. I don't put my real MVS userid on the Internet.

When are "Flower Box" output created?

IBM only gives the SMF exit IEFACTRT the ability to put messages in the SYSMSG dataset. This exit is driven when SMF 30 records are written at job step and end termination.

How do I link in the IBM SYSMSG routine needed by an IEFACTRT routine? [2006-01-01]

You can put these instructions at the very bottom of your IEFACTRT routine:

 PUNCH ' ENTRY   XYZACTRT'                                            
 PUNCH ' INCLUDE AOSB3(IEFTB724) '                                    
 PUNCH ' NAME    XYZACTRT(R)'                                         
         END ,                                           
 

And then have a DD card in your link step: //AOSB3 DD DISP=SHR,DSN=SYS1.AOSB3

Hey, doesn't other SMF records drive IEFACTRT also?

Okay SMF4 and SMF5 records can drive IEFACTRT but they are obsolete. IBM recommends that you use SMF30 records instead of the old 4 and 5. All new enhancements are only put in SMF30 records. Your shop probably follows IBM recommendations and doesn't write SMF4 or SMF5 records anymore.

How do test a new IEFACTRT exit without IPLing? [2006-01-01]

Here is an example of dynamically adding a IEFACTRT routine named IPOACTRT with the IBM SETPROG commands for just jobs that start with "IBMUSER":

setprog exit,add,exitname=sys.iefactrt,modname=ipoactrt,dsname=
ibmuser.dave.load,jobname=ibmuser*        

See this example of testing Rob Scott's IEFACTRT routine.

How do I dynamically replace my IEFACTRT exit without IPLing?

Here is an example of dynamically replacing a IEFACTRT routine named XYZACTRT with the IBM SETPROG commands:

SETPROG EXIT,DELETE,EXITNAME=SYS.IEFACTRT,MODNAME=XYZACTRT

SETPROG EXIT,ADD,EXITNAME=SYS.IEFACTRT,MODNAME=XYZACTRT,DSNAME=SYS2.XYZ.LINKLIB

Where you issued the command for SYS.IEFACTRT, you may need to also do SYSJES2.IEFACTRT, SYSTSU.IEFACTRT and/or SYSSTC.IEFACTRT (if the IEFACTRT exit was originally defined that way via the SETPROG or PROGxx member).

You might recall that Omegamon use to have a SMF exit replace function. Instead of taking some time to use the IBM API to do this easier than the IBM SETPROG commands, they just yanked that function out of Omegamon.

I want to send you my Flower Box output. How can I help you get it ready for the web?

You can:

• Only send text files. Please no word processing documents or graphic files
• Please change your userid (if in the output) to "IBMUSER" or disguise it as you see fit.
• Please change your SMF id and NJE to something like XYZ or whatever. I think that it's best to not post corporate internal network system names on the Internet. Too many hackers, not enough Jolt.

What do you assume when you get my Flower Box output?

Unless you specify otherwise:

• If you state that the output is not for the web, I will keep for my private collection.
• If you don't state any restrictions, I may put it on my web page.
• If you like, send me anonymous contributions and I'll put them here without implicating you. I want to see what you're shop does. Really please send it...

That's a strange hobby you got Dave. Are you seeking treatment?

The first step to recovery is admiting you have a problem. I'm telling the world: send me your poor, your huddled masses, your IEFACTRT-created Flower Boxes.

Seriously, I'm interested in the visual display of information. I like to see what other SMF record fields are displayed and to compare my output and code.

XYZ120I ------------------------------------------------------------------------ S t e p E n d S t a t i s t i c s Step Name: SMPE Cond Code: 0004 Start: 01-Jul-1999 10:22:00 AM Step Num: 1 PGM Name: GIMSMP End: 01-Jul-1999 10:27:37 AM CPU (TCB): 00:00:48.44 Storage below 16M: 4,032k CPU (SRB): 00:00:03.92 Storage above 16M: 32,444k Trans Act: 00:05:34.73 Service Units: 1,884,723 Tape Mnts: 0 Total EXCPs: 29,404 Unit-- DDName-- EXCP Count-- Blksize Unit-- DDName-- EXCP Count-- Blksize 0EC3 D SYSUT1 60 3,200 0E29 D SYSUT2 4,344 0 0E29 D SMPWRK1 17 23,440 0EC3 D SMPWRK3 1,677 3,200 0EC3 D SMPWRK6 12 23,440 0C10 D SMPSCDS 1 3,120 0C15 D MIGLIB 1,972 32,760 0C15 D LINKLIB 5,424 32,760 0C15 D LPALIB 1,784 32,760 0C15 D CSSLIB 615 32,760 0C15 D MACLIB 2,424 6,160 0C15 D PARMLIB 77 6,160 0C15 D PROCLIB 76 6,160 0C15 D SAMPLIB 1,353 3,120 0C10 D SMPCSI 504 0 0C10 D MVST100 6,645 0 0CBC D SMPPTS 476 8,800 Total DASD EXCPs: 27,461 Total Tape EXCPs: 0 ---------------Paging--------------- ---------------Paging--------------- Common (LPA+CSA): 7 Hiperspace: 0 XYZ121I ------------------------------------------------------------------------ IEF373I STEP/SMPE /START 1999182.1022 IEF374I STEP/SMPE /STOP 1999182.1027 CPU 0MIN 48.44SEC SRB 0MIN 03.92S

XYZ120I ------------------------------------------------------------------------ J o b E n d S t a t i s t i c s Job Name: IBMUSERT System ID: XYZ1 Submit: 01-Jul-1999 10:21:05 AM Job JESID: JOB10960 MVS level: SP6.0.6 Start: 01-Jul-1999 10:22:00 AM Job Class: Q End: 01-Jul-1999 10:27:37 AM CPU (TCB): 00:00:48.44 Storage below 16M: 4,032k CPU (SRB): 00:00:03.92 Storage above 16M: 32,444k Trans Act: 00:05:34.73 Service Units: 1,884,723 Tape Mnts: 0 Total EXCPs: 29,404 ---------------Paging--------------- ---------------Paging--------------- VIO In: 0 VIO Out: 0 Pages In: 0 Pages Out: 0 Common (LPA+CSA): 7 Hiperspace: 0 VIO Reclaims: 0 Pages Stolen: 0 ACF2 User: IBMUSER Programmer: ALCOCK, DAVID Acct Info: D,D,D,SYS XYZ122I ------------------------------------------------------------------------ IEF375I JOB/IBMUSERT/START 1999182.1022 IEF376I JOB/IBMUSERT/STOP 1999182.1027 CPU 0MIN 48.44SEC SRB 0MIN 03.92S

Ramblings:

• I put a message ID of "XYZ120I" at the start of each Flower Box. This allows the SDSF user to quick go through them with quick "FIND XYZ120" commands.
• The format of the I/O by DDNAME entries created at Step Termination are "uuuu t dddddddd" where:
  • "uuuu" is the UCB
  • "t" is the type: D for DASD and T for Tape.
  • "dddddddd" is the DDName

Source:

J E S 2 J O B L O G -- S Y S T E M P 3 9 0 -- N O D E N 1 ... IEF403I IBMUSERG - STARTED - TIME=13.13.13 -Jobname Stepname Procstep RC EXCP Conn TCB SRB Clock Serv PG Page Swap VIO Swaps -IBMUSERG DOIT STEP1 00 3036 0 .04 .01 .3 39763 1 0 0 0 0 IEF404I IBMUSERG - ENDED - TIME=13.13.32 -IBMUSERG Ended. Name-David Alcock Total TCB CPU time= .04 Total Elasped time= .3

J E S 2 J O B L O G -- S Y S T E M P 3 9 0 -- N O D E N 1 ... IEF403I IBMUSERG - STARTED - TIME=13.13.13 XYZ101I Jobname Stepname Procstep RC EXCP Conn TCB SRB Clock Serv PG Page Swap VIO Swaps XYZ102I IBMUSERG DOIT STEP1 00 3036 0 .04 .01 .3 39763 1 0 0 0 0 IEF404I IBMUSERG - ENDED - TIME=13.13.32 XYZ103I IBMUSERG Ended. Name-David Alcock Total TCB CPU time= .04 Total Elasped time= .3 Now they also line up better with the IEF403I/IEF404I messages.

EXIT ADD EXITNAME(SYS.IEFACTRT) MODNAME(IEEACTRT)	IBM's sample IEFACTRT
EXIT ADD EXITNAME(SYSJES2.IEFACTRT) MODNAME(IEEACTRT)	IBM's sample IEFACTRT
EXIT ADD EXITNAME(SYS.IEFACTRT) MODNAME(XYZACTRT)	My IEFACTRT routine (creates flower boxes)
EXIT ADD EXITNAME(SYSJES2.IEFACTRT) MODNAME(XYZACTRT)	My IEFACTRT routine (creates flower boxes)

• Tom Lewis sent me his IEFACTRT-created flower box output. He said "This IEFACTRT has been upgraded many times over the past 25+ years!". [2009-06-18]
• Kevin Keith sent me this IEFACTRT output.
• Rob Scott (of MXI fame) (includes howto for testing)
• IBM Class IEFACTRT routine (IBM parallel sysplex class in Orlando I think)
• MVT IEFACTRT output I'm not sure about who created this IEFACTRT routine but it's included with the MVT available from Yet another MVT distribution.
• Washington State University - from a post to IBM-Main on 1999-09-02
• From when I worked at CSW circa 1994 - Created by the Kommand SMF exit
• From when I worked at Rockwell International - I didn't create this output.
• From Neimans - My second attempt at Flower Boxes. I quickly got tired of the IDCAMS-like display of information.
• A friend sent me this - he unwittingly was making a contribution to this page.
• John Ferguson sent me this
• David G T Rawson sent me this
• Kevin Olive sent me this
• Obtained from some IBMLINK function - Created output from some nameless IBM data center.
• From 1993 CBT tape files 109 and 110 - I don't know if this is still there. I plan to put pointers to places on the http://www.cbttape.org of sample IEFACTRT exits that reside there.