|
View previous topic :: View next topic
|
| Author |
Message |
Akatsukami
Global Moderator
Joined: 03 Oct 2009
Posts: 1788
Location: Bloomington, IL
|
|
|
|
Can't be done with vanilla MVS/JES2/JES3, I agree.
But suppose I were to write my own DSP to replace the JCL converter/interpreter (not that I plan on trying  ). Could the JCL at least be expanded to allow multiple executions of a step? Could we go further and introduce a DO WHILE/UNTIL construct with, perhaps, other DSP mods (such as MAIN)? |
|
| Back to top |
|
 |
William Thompson
Global Moderator
Joined: 18 Nov 2006
Posts: 3156
Location: Tucson AZ
|
|
| Back to top |
|
|
Robert Sample
Global Moderator
Joined: 06 Jun 2008
Posts: 8696
Location: Dubuque, Iowa, USA
|
|
|
|
| If you write your own converter / interpreter, you can have it do pretty much anything you want. There would be design issues, of course (such as backwards compatibility with current JCL) -- but if you've got the skill to write the software then you can have it doing loops, true IF/THEN/ELSE logic, arithmetic, and so forth. |
|
| Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006
Posts: 6250
Location: Mumbai, India
|
|
|
|
This thread smells ambiguous ...  |
|
| Back to top |
|
|
Willem Vermeer
New User
Joined: 31 Oct 2007
Posts: 38
Location: Amsterdam, the Netherlands
|
|
|
|
I think that introducing such a loop-capability would lead to very serious incompatiilities with any current JES-environment. After all, what would you do with, say, DISP=(,PASS)? Which one would you be PASSing? Or with a DISP=(,CATLG)? Try executing that multiple times!
What I would like to see, would be a facility to execute Jobsteps concurrently. If the steps have no interdependencies, why wait for eachother? |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix
|
|
|
|
Hello,
| Quote: |
| If the steps have no interdependencies, why wait for eachother? |
One reason could be performance issues. . . |
|
| Back to top |
|
|
Willem Vermeer
New User
Joined: 31 Oct 2007
Posts: 38
Location: Amsterdam, the Netherlands
|
|
|
|
True, any resource can be exhausted but I don't consider this to be a problem that's fundamentally different from what I would be facing with just 1 step.
And look at DFSMSdss. This utility allows you to execute several operations in parallel. Indeed, the manual warns you about availability of resources, but so what?
I would really like to see JES2 do something similar. |
|
| Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix
|
|
|
|
Hello,
| Quote: |
| True, any resource can be exhausted but I don't consider this to be a problem that's fundamentally different from what I would be facing with just 1 step. |
On multiple systems i've been involved with someone asked me about this or that "tuning concern". . .
What happened in several cases was that processes that ran concurrently ran far longer than when run serially. The culprit was that more time was spent moving the heads on the dasd than actually processing the data. Of course, this was some time ago when the dasd was real 33nn disk and not emulated. . .  |
|
| Back to top |
|
|
|
|
