FWIW The master catalog should be used only to point hlq's at user
catalogs and not to catalog fully qualified DSNs - for performance
reasons. E.g. the hlq "JOEBLOW" should be catalog'd as a pointer to a
usercat and should not be recorded as the fully qualified DSN mcat entry
"JOEBLOW.MY.FOO.BAR". Otherwise the mcat becomes overloaded with fully
qualified DSNs - all of which have to be checked for each DSN
allocation, before (if no mcat entry found for it) transferring the
search to the mcat's pointed-at usercat ... thus degrading overall
system performance. The fewer mcat entries there are, the better. Run a
"LISTCAT ENT('<master catalog name>'),CATALOG" to see how much junk is
stored in the master catalog that has to be checked at each dataset's
allocation. My ha'pennyworth - CP
Jesse 1 Robinson wrote:
This elementary question has been promoted to at least middle if not high school. I for one am still itching to learn why it's deemed worthwhile. In any case, I suggest putting a script in place to simplify (re)specification. The sample below can be adapted to the user's environment.
It's assumed that the SMPE release HLQ is OSRnn, where nn is the release id
like 13 or 21. DSN and VOL can be specified, or the Rexx can be typed on a 3.4
display that shows the data set on the volume. The maintenance user catalog is
likewise 'MVSRnn.ICF.MASTER'.
/* REXX */
ARG dsn vol
IF vol = "" THEN DO /* assume ISPF 3.4 */
ADDRESS ISPEXEC "VGET (ZDLVOL)" /* avail on ISPF 3.4 screen */
vol = ZDLVOL
END
dsn = STRIP(dsn,B,"'")
TRACE C
"DEL '"dsn"' NSCR" ,
"CAT('MVSRnn.ICF.MASTER')"
"DEF NVSAM (NAME('"dsn"') DEVT(3390) VOL("vol"))" ,
"CAT('MVSRnn.ICF.MASTER')"
"DEF ALIAS (NAME('OSRnn."dsn"') REL('"dsn"'))" ,
"CAT('MVSRnn.ICF.MASTER')"
.
.
J.O.Skip Robinson
Southern California Edison Company
Electric Dragon Team Paddler
SHARE MVS Program Co-Manager
323-715-0595 Mobile
626-302-7535 Office
[email protected]
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf
Of John Eells
Sent: Tuesday, August 30, 2016 10:49 AM
To: [email protected]
Subject: (External):Re: Elementary dataset alias question
Tom Marchant wrote:
<snip>
A word of caution. If the real data set is subject to moving to a
different volume, for example because of HSM migrate/recall, the
second catalog entry will not be updated when it is recalled. And of
course, if the data set is on the IPL volume, you would catalog it to ****** or the appropriate symbol.
<snip>
Tom raises two important points.
To widen the aperture on his first one a bit further, it is of course not only
DFSMShsm that can move things around. For example, whoever maintains the data
sets that are aliased this way must always remain conscious of the additional
catalog entries that are perhaps maintained in different catalogs than the
primary entries, and change them in concert in the correct catalog(s) when and
if data sets move between volumes.
--
John Eells
IBM Poughkeepsie
[email protected]
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
