Hi List,

Thanks to those of you who responded. As Mark pointed out, I could install into 
a program controlled TCP/IP library with a USERMOD and perhaps I should. If IBM 
distributed TCP/IP with the MOD or LMOD (i.e. FTCHKCMD) defined, I would be 
obligated to install with a USERMOD. Since FTCHKCMD is unknown to SMP/E, I 
decided to install into our "local" LINKLIB. 

It seems to me, however, that this issue could arise with a software component 
that is not part of z/OS. As a general rule, our non-z/OS target DSNs (i.e. 
true names) include a version-release-level qualifier. Installation and 
maintenance is done using these DSNs (which are specified in the DDDEFs). They 
are accessed by executables (e.g. JCL and EXECs) via a SYMBOLICRELATE alias 
that does not incorporate the version-release-level qualifier (which is 
represented in the SYMBOLICRELATE with a symbol).

For example, DSN=SYS2.LINKLIB.Z19 has a SYMBOLICRELATE alias of 
SYS2.LINKLIB&MVSDSNQ (the symbol &MVSDSNQ has a value of .Z19 for z/OS 1.9). 
This method allows us to upgrade or revert many non-z/OS products simply by 
changing the value of a symbol (with a few caveats that I won't discuss here).

It irritated me that RALTER ADDMEM was not cooperating. I found a work-around. 
I admit that it's not pretty and I'm not recommending it but it does work. We 
have an automated startup process (after IPL). The following command is now 
issued after the RACF subsystem initializes:

        #RALTER  PROGRAM  FT*  ADDMEM('SYS2.LINKLIB&MVSDSNQ'//PADCHK)

Because it's a command processed by CONSOLE, symbolic substitution occurs:

        IEE295I COMMAND CHANGED BY SYMBOLIC SUBSTITUTION                      
        ORIGINAL: #RALTER  PROGRAM FT* ADDMEM('SYS2.LINKLIB&MVSDSNQ'//PADCHK)   
                                                                
        MODIFIED: #RALTER  PROGRAM FT* ADDMEM('SYS2.LINKLIB.Z19'//PADCHK)

If RALTER processing finds that the specified DSN is already defined, it 
ignores the request (i.e. we don't end up with multiple occurrences of the same 
DSN). Obviously, as the value of the symbol changes (due to upgrades), the 
outdated DSN entries will remain undeleted.

Thought that some of you may find this interesting....

Alan   
 



 

-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of 
Mark Zelden
Sent: Tuesday, July 13, 2010 05:47
To: [email protected]
Subject: Re: RACF PROGRAM ADDMEM and SYMBOLICRELATE

On Mon, 12 Jul 2010 15:52:29 -0700, Starr, Alan <[email protected]>
wrote:

>Hi list,
>
>I just added an FTP exit to a user loadlib that has a DSN (VTOC entry) 
>of
SYS2.LINKLIB.Z19. This dataset is assigned a SYMBOLICRELATE alias of 
SYS2.LINKLIB to facilitate z/OS-release-independent accesses.
>
>FTP Server requires access to the FACILITY called BPX.DAEMON which, in
turn, requires that its exits reside in a "controlled" dataset.
>
>I therefore want to RDEFINE   PROGRAM   FT*   ADDMEM(dataset-name//PADCHK)
>
>I strongly suspect (but have not tested) that specifying the alias name 
>of
SYS2.LINKLIB is not going to produce the desired effect because the OPENed DSN 
(in LNKLST) is SYS2.LINKLIB.Z19 (i.e. the alias was resolved by OPEN and then 
forgotten).
>
>I have tried specifying a static system symbol (&MVSDSNQ = .Z19) as 
>part of
the DSN - ADDMEM('SYS2.LINKLIB&MVSDSNQ'//PADCHK) - but RACF didn't accept that.
>
>Am I stuck having to specify a full-qualified "truename" and change the
MEMBER segments of all PROGRAM resources every time I upgrade?
>

Yes, if you want to keep using that methodology.   Since there should
already be a PROGRAM "*" entry with plenty of "system" libraries specified, why 
not just create a USERMOD and link it into hlq.SEZALOAD or you could use 
SYS1.LINKLIB (but SEZALOAD seems more appropriate).

Mark
--
Mark Zelden - Zelden Consulting Services - z/OS, OS/390 and MVS       
mailto:[email protected]                                          
Mark's MVS Utilities: http://home.flash.net/~mzelden/mvsutil.html
Systems Programming expert at http://expertanswercenter.techtarget.com/

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions, send email to 
[email protected] with the message: GET IBM-MAIN INFO Search the archives at 
http://bama.ua.edu/archives/ibm-main.html

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to