iCan Blog Archive

I wrote about workload groups some time ago in Manage Workloads Better with IBM i 7.1.  The information in that older article is still accurate, and I’m not going to repeat a lot of what I wrote there.  If you have not yet used workload groups, you may want to read the older blog for additional information on why workload groups are useful. 

Workload groups allow you to limit the processor resources a workload is allowed to use.  A workload can be a single job, or it can be all jobs within a subsystem.  You create a workload group with the “Add Workload Group” (ADDWLCGRP) command.  It’s a very simple command – you simply give your workload group a name and specify the number of processors that workload is allowed to use.

You then have to associate the work with that workload group.  And this is where IBM i 7.3 has been enhanced from prior releases.  On the 7.1 and 7.2 releases, you had to create a data area that associated the workload group with the subsystem in which the workload runs.  The data area, QSYS/QWTWLCGRP, specified pairs of names – the subsystem name and the workload group name.  The contents of the data area were only checked when the subsystem was started, so you had to end, then start the subsystem for the configuration to take effect.

Here’s an example of creating the data area on the 7.1 or 7.2 releases:


Now with 7.3, the create and change subsystem commands have been enhanced with the “Workload Group” (WLCGRP) parameter, which has a default value of *NONE.  This new parameter allows you to specify the workload group name on the subsystem description.  If you use the change subsystem command to specify the workload group to use, the change is effective immediately.  Message CPI146C, Subsystem &1 is using workload group &2, is logged to the job log of the subsystem monitor job to reflect the workload group is in effect.

If you have been using workload groups on 7.1 or 7.2, and upgrade to 7.3, you will need to change your subsystem description to use the workload group as the data area is obsolete with the 7.3 release and is ignored.

This blog post was originally published on IBMSystemsMag.com and is reproduced here by permission of IBM Systems Media.