I had two reasons for suppressing superfluous NOTIFY attempts. -- Several junk messages in syslog each time about trying to find the user log, failing, and discarding the message. Over and over for the same handful of userids hard-coded into countless jobs.
-- Some obvious though probably trivial overhead of checking for a data set that does not now and most likely never will exist. I did not want to substitute my own overhead. (Note that RACF would not know whether the user log exists; only the catalog.) So I just built a small table of known problem userids. Scan the table. If it matches, put out one line of doc. All without the rigmarole of trying to change maybe thousands of job streams that have been around for decades. $HASP954 Exit 16 suppressed NOTIFY for xxxxxxx . . J.O.Skip Robinson Southern California Edison Company Electric Dragon Team Paddler SHARE MVS Program Co-Manager 323-715-0595 Mobile 626-543-6132 Office ⇐=== NEW [email protected] -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of John McKown Sent: Thursday, June 29, 2017 10:52 AM To: [email protected] Subject: (External):Re: JES2 Exit 16 (was Re: how to keep messages in the sys1.broadcast for ever) On Thu, Jun 29, 2017 at 12:34 PM, Paul Gilmartin < [email protected]> wrote: > On Thu, 29 Jun 2017 12:12:50 -0500, John McKown wrote: > > >On Thu, Jun 29, 2017 at 12:07 PM, Paul Gilmartin wrote: > > > >> Are entries in SYS1.BRODCAST timestamped so they could be aged out? > >No. > > > >This originated back is MVT days. MVT did not really have an API for > >dynamic allocation. The data set was "fake opened" (i.e. code other > >than OPEN found the DSN & extents, and built an "open" DCB and > >associated > DEB.) > >The DCB was in common storage so that the SEND command could just > >write to the DSN using it. Much like what CA-1 does with the TMC. > > > Seems like an integrity nightmare. > It was designed long before IBM issued any of their security assurances. Also, I am saying that is how it was done WAY BACK WHEN. I don't know how it is done now. I would __guess__ (SWAG), with individual TSO broadcast data sets, that the SEND command now uses standard DYNALLOC / OPEN and so on. > > It's past overdue to trade it for a dog, then shoot the dog. > Dog has been upgraded to Cerberus quality. > > >> UNIX facilities could be your friend here, even as UNIX batch jobs > simply > >> append their logs to the user's mbox. > > -- gil > -- Veni, Vidi, VISA: I came, I saw, I did a little shopping. Maranatha! <>< John McKown ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
