Regularly. Disadvantage: you have to create it. OK that's three
minutes and about five commands. Advantage: Call it what you will, and
if you have other things you want the controlling subsystem to do you
can do that without destroying the IBM supplied subsystem. At upgrade
time it also avoids having to redo any customizations you made to the
IBM supplied subsystem since it will not be over written by IBM.
If you really want to, you can just create a duplicate of QCTL and call
it what you want as well.
Word of caution: It's best to create a library to hold all of your
custom system level objects in it (such as a custom controlling
subsystem, job descriptions, classes, routing programs etc...). I
usually call the libraries PARTITION for partition level objects and
SERVER for objects that are common to all partitions on the system.
I'll change the system library list to have PARTITION and SERVER before
QSYS at the top of the library list so your objects get picked up before
the IBM objects do if you happen to name them the same. While your
custom objects could go into QGPL or QUSRSYS why not keep those
libraries clean and create your own?
Remember at when doing an upgrade or initial install there are a couple
of other IBM supplied subsystems that are temporarily used as
controlling subsystems. DO NOT mess with them. First off there's no
need and secondly you'll find installation processes don't run as expected.
Chief Technical Architect
Agile Technology Architects
On 8/29/2013 6:55 AM, rob@xxxxxxxxx wrote:
Well, I guess there's probably an infinite number of possibilities for
controlling subsystems. Well, not really. It's probably limited to the
number of legitimate subsystem names you can create with 10 characters or
Anyone use anything different for a controlling subsystem than QBASE or