Today we dwell into one of the exciting topic in the mainframe world its “Capping”.
What is capping?
Capping controls the amount of CPU consumption by an IBM z System logical partition LPAR. In this context, CPU here is general purpose CPU and not a specialty processors like z Integrated Information Processor (zIIP), Integrated Facility for Linux (IFL) or Internal Coupling Facility (ICF)
Why a need for capping?
There could be several reasons. One of the most compelling reason for a need of capping is to control cost. There are pricing models that bills based on the CPU usages. Work on mainframe computers is often measured in millions of service units (MSUs), which is a measure of the processor (CPU) capacity used to execute the work. The mainframe operating system generates monthly reports that determine the customer’s system usage (in MSUs) during every hour of the previous month using a rolling average (e.g., a 4-hour rolling average) recorded by each LPAR or a capacity group for the customer. To control costs, user might assign each LPAR or capacity group a consumption limit (Defined Capacity or Group Capacity Limit), and it cannot use more MSUs than allotted in its respective consumption limit. Workload Manager (WLM) controls this kind of capping.
What capping options does mainframe provide?
It is best to put it in a table. Capping is controlled either by PR/SM or WLM.
|Capping Type||Controlled by||Where to specify|
|Group Absolute Capping||PR/SM||HMC|
|Absolute MSU Capping||WLM||HMC, IEAOPT|
|Resource group Capping||WLM||WLM Policy|
In an another post we will go into each of these capping options and how to know if these capping are enforced i.e. where to look if you need to know if an LPAR is capped.