Re: SAVRST command fun...

Folks:

V5R4 CLLE program.

With further testing it turns out that with SAVRST at V5R4 when you use the CHGPERIOD keyword, the only value that winds up being accepted is the start date:

Time period for last change: CHGPERIOD
Start date . . . . . . . . . . *ALL
Start time . . . . . . . . . . *ALL
End date . . . . . . . . . . . *ALL
End time . . . . . . . . . . . *ALL

The other values cause an error when called from a CL program (submit too from CL to answer Joe's question) When called from the command line, you can use them all.

This behavior changes at V7R1 (I don't have a V6 system I'm willing to test with right now) where it works properly with all the values specified from a program.

As to the '/' in QSYS.LIB, the command is smart enough to put it in for you if you don't specify it, and happily goes along when you do. I saw that and corrected it but it made no difference in the way the program ran.

I don't know if there is a PTF that fixes it or not, I have not dug into it that far.

Jim Oberholtzer
Chief Technical Architect
Agile Technology Architects


On 12/29/2011 4:26 PM, Joe Pluta wrote:

On 12/29/2011 4:04 PM, Scott Klement wrote:

> hi Jim,
>
> This is just a guess, I've never used SAVRST. But...
>
> Could it be because you're missing the leading slash on /QSYS.LIB? So
> it's going to use a different file system, depending on what the CWD of
> the job is?
>

Yup that sounds about right. It would explain the difference between
two different jobs. It's not so much a matter of how it's being called,
but what the current environment is.

In fact, I'll bet that when it works the CWD is / and so the QSYS system
is being omitted, but on the program call the CWD is something other
than /, and so the *OMIT for QSYS.LIB isn't matching anything and the
command is attempting to save QSYS.LIB (and thus the CPF382B).

Joe
--