2. PERFORMANCE REPORTING ANALYSIS › 2.1 ILM Analysis › 2.1.2 Usage Guidelines › 2.1.2.2 Estimating MSU Consumption for non-z/OS Workloads

2.1.2.2 Estimating MSU Consumption for non-z/OS Workloads

The Estimated MSU Consumption Rate Report analyzes historical
HARCPU data collected from non-z/OS workloads, calculates the
max MSU consumption rate, the average MSU consumption rate,
and the MSU four-hour rolling average.  These estimations are
accomplished by extracting the processor model and the number
of processors from the historical data, and pulling the
associated MSU conversion factor from a provided table.  For
further discussion on how this MSU conversion factor is used,
see section 2.1.3.2.

When analyzing the report generated by this query here
are some guidelines to follow:

o Setting your defined capacity

  Ideally, your peak MSU consumption rate should always be as
  close as possible to the defined capacity, without
  exceeding it. A general rule to follow when setting your
  defined capacity is set it near the MSU consumption rate's
  95th percentile.

  Remember capping impacts all workloads running in the
  partition and can result in elongated response times and
  generally unacceptable performance. The following are some
  additional guidelines to follow once you have set the
  defined capacity.

   o If the peak MSU consumption rate is constantly lower
     than the defined limit, you may consider decreasing the
     LPAR capacity until WLM begins capping the partition (a
     capping percentage of 1% to 5% ensures the work is done
     optimally at the lowest cost).

   o If the peak value exceeds the defined capacity for long
     periods (high WLM capping percentage reported), you may
     consider increasing the LPAR capacity limit until the
     WLM capping percentage falls to within an acceptable
     range.

   o If the peak value occasionally exceeds the defined
     capacity (low WLM capping percentage), you may consider
     time shifting workloads.

o Estimated MSU Four-Hour Rolling Average values are zero

  Due to the nature of a rolling average, you need to process
  a certain amount of data before you can begin calculating
  the rolling average.  In the case of the Estimating MSU
  Consumption Rate report the MSU four-hour rolling average
  will be zero until you have processed four hours worth of
  data.

  This becomes important to remember if you decide to report
  on an individual zone. If the zone consist of five hours,
  the report will display zero's for the MSU Four-Hour
  Rolling Average for the first four hours and only give you
  values for the last hour.

o Warning message PER00121W is issued in the MICSLOG

  Warning message PER00121W is issued when the historical
  input data contains combination of processor model and
  number of processors not defined in the MSU Conversion
  Table. If this message is issued update, the MSU Conversion
  Table through the extended options panel.  To update this
  table you will be required to have the processor model, the
  number of processors, and MSU rate associated to it.