SMF 30’s

Sent from my iPhone

No one said I could type with one thumb 

> On Nov 9, 2023, at 18:34, Lennie Dymoke-Bradshaw 
> <[email protected]> wrote:
> 
> I believe the capability of understanding and counting program LOADs is in 
> the latest version of SDSF for z/OS 3.1. (I hope Rob Scott will correct this 
> if I am wrong).
> However, I do not think this necessarily answers the question posed. That 
> question related to the number of times a program is executed, rather than 
> the number of times it is LOADed, LINKed to or even ATTACHed. A program can 
> be loaded (using a LOAD SVC) and then executed multiple times. That execution 
> can be via a LINK SVC but could just as easily be via a CALL, which is 
> effectively a BASR or BALR, a machine instruction which does not offer the 
> level of traceability that the LOAD, LINK and ATTACH services offer. As such 
> a load module monitor such as that in SDSF will not address the issue.
> If the load module is marked not reusable and not reentrant, then I think it 
> is unlikely to be reused after a first execution. I would expect it to be 
> DELETEd and then re-LOADed. I don't think normal processing of the module 
> using language environment will allow reuse.
> If that is the case, then the question might be able to be answered for a 
> specific module that is neither reentrant nor reusable.
> Lennie Dymoke-Bradshaw
> https://rsclweb.com 
> ‘Dance like no one is watching. Encrypt like everyone is.’
> 
> -----Original Message-----
> From: IBM Mainframe Discussion List <[email protected]> On Behalf Of 
> Steve Thompson
> Sent: 09 November 2023 22:15
> To: [email protected]
> Subject: Re: SMF record for number of program executions?
> 
> If you are willing to write an exit to get the info, you can get it via a CSV 
> exit (I forget its name, but ALL "LOAD"s go through it). Understand, if you 
> use that exit, it has to have a very short code path, can't cause a wait of 
> any kind, or you will cause problems for all address spaces in that LPAR. The 
> idea is to capture the DSN & member and immediately write it to an SMF buffer 
> or similar so you can immediately return control.
> 
> But other than what others have said, there is no other way to see all 
> dynamically loaded subroutines or load-modules. You will not capture static 
> routines as the LNKEDT doesn't use that interface.
> 
> I believe that IBM Products make use of that or another undocumented path 
> through VLF that is handling LLA and a bit of caching of modules.
> 
> Regards,
> Steve Thompson
> 
> 
> 
>> On 11/9/2023 4:56 PM, Glenn Miller wrote:
>> Hi Linda,
>> When I have been requested to provide that information, I have used the IBM 
>> Z Software Asset Management ( aka iZSAM ) software product, which was 
>> previously known as IBM Tivoli Asset Discovery for z/OS ( aka TADz ).
>> 
>> Glenn Miller
>> 
>> ----------------------------------------------------------------------
>> 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
> 
> ----------------------------------------------------------------------
> 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

Reply via email to