We are at V7R2M0 on all referenced IBMi's.
I've a process (RPG with imbedded SQL) that will read data from a
remote IBMi DB, process it, persist it to the local IBMi DB, delete
the remote row and move to the next in a loop.
This thing works just fine in the development environments. When the
need for an SQLPKG arises, the package is automatically created on
the remote as needed. The program object is created using a profile
that is available on both the remote and local machines. As weird as
this seems to me, this will become important.
As we move to the next level of testing, the creator is that of the
move coordinator, who does not have a valid user profile on the
remote machine. This seems to be causing an issue.
As the programs runs, the need to create the SQLPKG on the remote
arises. Rather than use the profile that is running the process as
the RDB user ID, it is retrieving the move coordinator profile that
created the program to create the SQLPKG. Because this profile does
not exist on the remote machine, the package creation fails. Which
causes the process to fail.
5770SS1 V7R2M0 140418 -Create SQL package- 09/16/15 09:25:05 Page 1
MSG ID ... TEXT
SQL0204 ... THE_MOVE_COORDINATOR in QSYS type *USRPRF not found.
SQL5056 ... SQL package creation for module PROGRAM_NAME failed.
Package name was to be PROGRAM_NAME in QTEMP at REMOTE_SYSTEM.
Not sure how to proceed.
While it appears that the SQL0204 error is the killer, I can't even
find SQL5056 on IBM's website.
I expect that I'll have to ask IBM about it, but thought I would
share this with you all.