[749] in Enterprise Print Delivery Team
Accounting and PSF/VM
daemon@ATHENA.MIT.EDU (Rocklyn Clarke)
Tue Dec 5 13:05:09 2000
Date: Tue, 05 Dec 00 12:59:37 EST
From: Rocklyn Clarke <RCLARKE@MIT.EDU>
To: Enterprise Print Delivery Team <PRINTDEL@MIT.EDU>
Cc: Karen Fortoul <fortoul@MITVMA.MIT.EDU>,
David G Kalenderian <davek@MITVMA.MIT.EDU>, Rachel Sage <sage@MIT.EDU>
Message-Id: <001205.130323.EST.RCLARKE@MIT.EDU>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Content-Transfer-Encoding: 7BIT
Hello Enterprise Printing Team Members,
As you know, we are currently trying to determine what features we should ask
IBM to incorporate into IPM/AIX accounting support. Towards this end we are
examining the way in which PSF/VM accounting works on the mainframe.
Yesterday I checked the accounting programs for VM and determined that we
do not currently interpret any of the flags set by PSF/VM accounting records
in a way which would indicate a failed or flushed print job. The FLUSH field
exists in the print file accounting tables mind you, but we do not set or
update it in any way based on the fields in the PSF/VM accounting records.
Line printer records are a different story, but they are not relevant to this
discussion.
I did notice however, that 4 of the 5 records produced by PSF/VM include a
processing code field in byte 52. The PSF/VM System Programming Guide has
the following to say about this field:
SFCM records '10'x and '40'x:
Byte 52
Contains the processing code.
The processing code indicates whether the file was successfully
processed by the SFCM. If the printing ended abnormally, this code
reflects the reason. The valid codes are:
0 Specifies that processing of the print file completed normally.
1 Specifies that processing of the print file ended due to user
error.
2 Specifies that processing of the print file ended due to system
error.
PDM records 'A0'x and 'D0'x:
Byte 52
Contains the processing code.
The processing code indicates whether the file was successfully
processed by the PDM. If the printing ended abnormally, this code
reflects the reason. The valid codes are:
0 Specifies that processing of the print file completed normally.
1 Specifies that processing of the print file ended due to user
error.
2 Specifies that processing of the print file ended due to system
error.
3 Specifies that processing of the print file ended due to operator
intervention.
Thus, the PSF/VM product does indeed provide information about print job
success or failure. We should ask IBM to do the same for IPM/AIX.
On the mainframe side, it is now clear that my accounting code is incomplete
and should be modified to indicate failed print files. Currently print files
which fail at the SFCM stage and do not make it to the PDM machine are not
charged. Unfortunately, it is likely that jobs which failed or were flushed
at the PDM stage, and generated the expected 2 accounting records (albeit with
the suitable error value in the process code) were charged to the customer
anyway. Oh well, more work for me.
Rocklyn
P.S. the PSF/VM system Programming Guide is available online at:
http://www.s390.ibm.com:80/bookmgr-cgi/bookmgr.cmd/Shelves/CD828106
Scroll to the bottom of the web page to find the Systems Programming Guide
