Re: Why can't a member of QPGMR change their password or run WRKSYSACT?

On 11-Sep-2015 21:38 -0600, Justin Dearing wrote:

<<SNIP>> why can't a programmer run WRKSYSACT? Its the equivalent of
taskmanager on windows or top on Linux. I know its a little bit of a
heavier command (and I'm ignorant of OS/400 internals to understand
why, but I accept it). However, I have access to WRKACTJOB,

Work With Active Jobs (WRKACTJOB) shows only jobs; drill-down is required for threads. The feature does not change the job that issues the request.

so I could write a program that does what WRKSYSACT does, but more
inefficiently.

Per what I noted above, the WRKACTJOB is quite minimal; while a program to mimic that feature is feasible, the Work With System Activity (WRKSYSACT) does much more, notably, showing both processes [aka jobs] *and* LIC tasks. I am guessing there is little chance being able to write the equivalent, inefficiently or otherwise.

Also, I could tie up the CPU through other methods as well.

The WRKSYSACT effectively changes the job issuing the command to have RUNPTY(0) or RUNPTY(-##) [though visibly manifest as Run Priority (RUNPTY) of 1, the real scale actually is -### to +###]. The feature was originally part of a purchased Licensed Program Product (LPP) [¿Performance Tools; PF1?], and because the tooling gave such an extreme priority over other work, the choice was to limit the request to just one job on the system.

Would seem reasonable to submit a request for change, perhaps asking that the feature be made available to secondary requesters. To increase the chance that the request would be accepted, perhaps suggesting to log a warning for secondary requesters [for the displayed output, perhaps also a visible indication manifest on the screen] that their request to WRKSYSACT will be run at the run priority of the job, or perhaps at a run priority of runpty+1 to minimize the impact; i.e. instead of failing the invocation with the error msg CPFC702 "Work with System Activity already active by user &4."