Re: PASTHR fails

Some comments inline below. Plus...

Ask of the other system [admin] for multiple device names which should have the appropriate Work Management setup, then specify multiple names on the VRTDEV() parameter, instead of limiting to just one name.

FWiW I recall that on our connections to Japan [& China] for passthru, we would always specify a VRTCTL() on which all of the USEnglish devices were setup; i.e. allocated to a subsystem with QSYS2924 or QSYS2938 as its SYSLIBLE() plus non-DBCS but compatible device types. The use of the named virtual controller avoided having to specify a list of device names for when one or more of those devices might already be in use or otherwise unavailable.

Burns, Bryan wrote:

The source system is in Japan and I don't have access to it.
It's a wonder I got the CPF translated.

Presumably the target system has a version of English from which an English version of the joblog can be spooled? As well even a spooled Japanese version will have formatting of its message identifiers same as in English spools; what the problem was in the JOB(x/y/z), in Japanese, might be sufficiently obvious with just the MSG#### logged. The text Rob included does not mention that also there is the possibility of the QAUTOVRT having enabled the Virtual Device Selection (QIBM_QPA_DEVSEL) exit point via *REGFAC or that the numeric limit is at fault [since v5r2 named devices on the QPACTL* and QVIRCD* virtual controllers are counted].

I'm auditing ZC entries of the devices so I know they're not
being used by someone else. Shouldn't they vary on
automatically?

A T-ZC would only occur if the device were actually changed, which would not necessarily be required and thus may not be logged. The device would change only when the model\type or other attribute needed to be changed to be compatible; perhaps when\if varied on.

Regards, Chuck

rob wrote:

Did you do a DSPMSGD CPF8916 and check every one of the
possibilities? Perhaps when it was in QVIRCD001 it might have
been:
- configured as something incompatible with the users terminal.
- varied off
- in use by someone else

Did you also check the joblog specified in the message on the
remote system?
Recovery . . . : Display the job log for job &5/&4/&3 at the
target system for messages that indicate which of the listed
errors occurred.

Burns, Bryan wrote:

User is getting CPF8916 "Cannot select virtual device DEV123 on
system S123456" when using the following in a CL:

STRPASTHR RMTLOCNAME(S123456) VRTDEV(DEV123) RMTUSER(JOHNDOE)
RMTPWD(DOEY1) but only if the device DEV123 is in controller
QVIRCD001.

Using the same command, if the device is in QPACTL01, pasthru
works normally.

What do I need to do to enable the device to be selected when
it's in QVIRCD001?