home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!paladin.american.edu!auvm!PENNDRLS.UPENN.EDU!TOM
- Return-Path: <TOM@PENNDRLS.UPENN.EDU>
- Message-ID: <9212171601.AA02893@noc2.dccs.upenn.edu>
- Newsgroups: bit.listserv.ibm-main
- Date: Thu, 17 Dec 1992 10:46:09 EST
- Sender: IBM Mainframe Discussion list <IBM-MAIN@RICEVM1.BITNET>
- From: "Thomas D. Denier" <TOM@PENNDRLS.UPENN.EDU>
- Subject: Re: Spool file size inconsistency
- Comments: To: IBM-MAIN%ricevm1.rice.edu@NOC2.DCCS.UPENN.EDU,
- IBMTCP-L%pucc.PRINCETON.EDU@NOC2.DCCS.UPENN.EDU
- Lines: 16
-
- I now understand why spool usage reported by the '$d spl,jobs=5' command was
- so far out of line with the record counts shown by SDSF for our TCPIP started
- task. All the SYSOUT data sets in the started task were specified as SYSOUT=*,
- causing them to inherit the spool class used for MSGCLASS. The MSGCLASS for
- started tasks was set to Z in the JES2 parameters, and class Z was set up for
- automatic purging in the JES2 parameters. For some reason the JESJCL and
- JESYSMSG data sets are not purged even though their spool class calls for
- purging. As John Pedlow and Steve Entwistle noted in responses sent to
- IBM-MAIN, the spool space used for purged data sets is not reclaimed until all
- of the SYSOUT data sets for a job are purged. In this case, that means that
- all the space occupied by the purged data sets from TCPIP is locked up until
- the JESJCL and JESYSMSG data sets are purged.
-
- The IBM-supplied JCL for the FTPSERVE started task also specifies SYSOUT=*
- for all SYSOUT data sets. I suspect that the same is true of all the other
- IBM-supplied JCL for started tasks assiciated with the TCP/IP product.
-