5.5.3. Configuring a Shared Processor Pool (Maximum Pool Capacity)

In order for a shared processor pool to be used, it must first be configured. At least the maximum pool capacity (max_pool_proc_units) must be set to an integer value greater than zero. A shared processor pool can be configured with the command “ms chprocpool” (change processor pool):

$ ms chprocpool ms11 SharedPool01 max_pool_proc_units=2

(The pool ID can also be specified instead of the pool name, which is often shorter.)

The generic pool name SharedPool01 is not very meaningful and should therefore be changed. It is best to choose a name that describes the purpose of the pool. We therefore rename the pool to testpool:

$ ms chprocpool ms11 SharedPool01 new_name=testpool

By specifying the maximum pool capacity, no processors or processor shares are reserved! The value of max_pool_proc_units can easily be greater than the number of available processors (the maximum possible value is 256), but the attribute then loses its meaning, since all LPARs together can at most fully utilize the number of available processors, but not more. The limitation by the configured maximum pool capacity can then never be reached in practice, the shared processor pool would thus be practically unlimited. Therefore max_pool_proc_units should always be smaller than the number of available processors.

The default shared processor pool DefaultPool is to be considered in particular. The attributes max_pool_proc_units and reserved_pool_proc_units cannot be set for this processor pool! The name DefaultPool cannot be changed either.