Re: (see thread on RPG list) Re: Test program calls puts(), creates spool file here, but not on customer box

On 21-Dec-2011 14:20 , James Lampert wrote:

CRPence wrote:

Something more agreeable would be to perform [a variation of] the
following override request before the test:

OVRPRTF FILE(*PRTF) DFRWRT(*YES) SPOOL(*YES) OUTQ(QGPL/QPRINT2)
MAXRCDS(*NOMAX) SCHEDULE(*JOBEND) HOLD(*YES) SAVE(*YES)
USRDTA(HELDTHIS) SPLFOWN(*JOB) SPLFNAME(*FILE)
USRDFNOPT(*NONE) USRDFNDTA(*NONE) USRDFNOBJ(*NONE)
WAITFILE(*CLS) OVRSCOPE(*JOB)

After the test... WRKJOB JOB() OPTION(*SPLF)

A DSPFD of both QPRINT and QSYSPRT shows SPOOL(*YES).

And with the above test, on the "problem" box, I still see no spool
file at all, not even in a FIN state, from the "host print" key, but
I *do* see it, with the USRDTA from the above override, on another
box, and if I do an Option 6 in PDM on the "problem" box, I *do* get
a spool file *with* the above override.

There were more parameters [since the v5r3 system where I composed the above] which might be necessary to have applied; e.g. the TOSTMF() someone else mentioned. I did not take the time to look for other parameters like that one, which might similarly hide the effect; Rob reported the TOSTMF() effect being no "FIN" spools tracked to the job, so that one is likely... presumably TOSTMF(*NONE) would best be added to the above invocation for the test.

Another issue for Host Print someone else had warned of, is the target telnet weirdness; i.e. final host destination for STRTCPTELN might redirect Host-Print requests, I think to the first telnet destination which is generally the emulated 5250 session. I seem to recall no matter what I tried to defeat that effect was for naught; noticed typically when signing on to customer's systems from the lab where an intermediate system was used for security.

I also have TRCJOB output (some 800-odd records; what am I looking
for?) from the scenario:
TRCJOB *ON
<host print key>
TRCJOB *OFF *OUTFILE

I have no access to TRCJOB on the system I use, so I can not analyze a functional request to do the same [nor do I know from what panel the request to Host-Print was requested] and then review to make a suggestion about what to look for. If you want, you can send the output [preferably from both a functional and a non-functional of the "same" request at the same OS level] and I will review that. Best to have the WRKJOB OUTPUT(*PRINT) taken for the same job, just before or just after the tracing.

Regards, Chuck