AS/400 Checking Page Fault Rates

Prerequisites

If there is data already in main memory, it can be referred to independent of the memory pool in which it is stored. However, if the necessary data is not available in any memory pool, it is brought into the same memory pool as the job that refers to it. (This is described as page faults.) When data is transferred into a memory pool, other data is moved, if it is being changed, and is automatically stored in the auxiliary memory. (This is described as paging.)

Procedure

Under OS/400: Pages in = Pages out.

Since page defaults hinder performance, the memory pool must be large enough to keep paging to an acceptable amount.

Value limits for page faults under OS/400

If in Transaction ST06 or on the CCMS Alert Monitor you notice a high page default rate in a memory pool, enlarge this pool.

Pool	Value Limit:
Machine pool (always system pool 1)	Since the machine pool contains objects used across the whole system, page faults in the machine pool affect all the jobs in the system. For this reason, ensure that the page fault rate is as low as possible in this pool. Value limit: 10 page faults per second
Other pools	For optimum performance, an average rate of 2 page faults per second for each active work process should not be exceeded. The upper limit that should not be exceeded is a rate of 20 page faults per second for each active work process. (see example below)

The following examples can help you when determining page fault rates in your system:

Best Performance:

Number of active work processes: (AS/400 jobs with status RUN)	10
Number of page faults for each active work process	2 faults per second
Permitted	20 page faults per second
Paging overhead for each work process (assumed average response time of the hard disk is 10 millisec.)	20 millisec. per second

This corresponds to a paging overhead of 2%.

Upper Limit

Number of active work processes: (AS/400 jobs with status RUN)	10
Number of page faults for each active work process	20 faults per second
Permitted	200 page faults per second
Paging overhead for each work process (assumed average response time of the hard disk is 10 millisec.)	200 millisec. per second

This corresponds to a paging overhead of 20%. The 20% limit should not be exceeded.